US20190371297A1 - Artificial intelligence apparatus and method for recognizing speech of user in consideration of user's application usage log - Google Patents
Artificial intelligence apparatus and method for recognizing speech of user in consideration of user's application usage log Download PDFInfo
- Publication number
- US20190371297A1 US20190371297A1 US16/539,773 US201916539773A US2019371297A1 US 20190371297 A1 US20190371297 A1 US 20190371297A1 US 201916539773 A US201916539773 A US 201916539773A US 2019371297 A1 US2019371297 A1 US 2019371297A1
- Authority
- US
- United States
- Prior art keywords
- user
- intention
- processor
- speech
- usage log
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
- 238000013473 artificial intelligence Methods 0.000 title claims abstract description 143
- 238000000034 method Methods 0.000 title claims description 36
- 230000005236 sound signal Effects 0.000 claims abstract description 20
- 239000013598 vector Substances 0.000 claims description 19
- 238000004891 communication Methods 0.000 claims description 12
- 238000013507 mapping Methods 0.000 claims description 7
- 238000005516 engineering process Methods 0.000 description 31
- 238000013528 artificial neural network Methods 0.000 description 28
- 230000006870 function Effects 0.000 description 26
- 230000003993 interaction Effects 0.000 description 14
- 238000010801 machine learning Methods 0.000 description 14
- 238000010586 diagram Methods 0.000 description 12
- 230000008569 process Effects 0.000 description 8
- 230000004044 response Effects 0.000 description 8
- 238000012549 training Methods 0.000 description 8
- 210000002569 neuron Anatomy 0.000 description 7
- 238000012545 processing Methods 0.000 description 6
- 238000013135 deep learning Methods 0.000 description 5
- 238000003058 natural language processing Methods 0.000 description 4
- 230000003287 optical effect Effects 0.000 description 4
- 238000004140 cleaning Methods 0.000 description 3
- 230000004913 activation Effects 0.000 description 2
- 230000006399 behavior Effects 0.000 description 2
- 238000002485 combustion reaction Methods 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 230000002787 reinforcement Effects 0.000 description 2
- 210000000225 synapse Anatomy 0.000 description 2
- 230000000946 synaptic effect Effects 0.000 description 2
- 230000000007 visual effect Effects 0.000 description 2
- 230000001133 acceleration Effects 0.000 description 1
- 230000003044 adaptive effect Effects 0.000 description 1
- 230000003190 augmentative effect Effects 0.000 description 1
- 230000008901 benefit Effects 0.000 description 1
- 239000003795 chemical substances by application Substances 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 230000001186 cumulative effect Effects 0.000 description 1
- 238000007405 data analysis Methods 0.000 description 1
- 238000013500 data storage Methods 0.000 description 1
- 230000007774 longterm Effects 0.000 description 1
- 238000010295 mobile communication Methods 0.000 description 1
- 238000007781 pre-processing Methods 0.000 description 1
- 239000011435 rock Substances 0.000 description 1
- 238000005406 washing Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
-
- G—PHYSICS
- G10—MUSICAL INSTRUMENTS; ACOUSTICS
- G10L—SPEECH ANALYSIS TECHNIQUES OR SPEECH SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING TECHNIQUES; SPEECH OR AUDIO CODING OR DECODING
- G10L15/00—Speech recognition
- G10L15/06—Creation of reference templates; Training of speech recognition systems, e.g. adaptation to the characteristics of the speaker's voice
- G10L15/063—Training
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F17/00—Digital computing or data processing equipment or methods, specially adapted for specific functions
- G06F17/10—Complex mathematical operations
- G06F17/16—Matrix or vector computation, e.g. matrix-matrix or matrix-vector multiplication, matrix factorization
-
- G—PHYSICS
- G10—MUSICAL INSTRUMENTS; ACOUSTICS
- G10L—SPEECH ANALYSIS TECHNIQUES OR SPEECH SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING TECHNIQUES; SPEECH OR AUDIO CODING OR DECODING
- G10L15/00—Speech recognition
- G10L15/02—Feature extraction for speech recognition; Selection of recognition unit
-
- G—PHYSICS
- G10—MUSICAL INSTRUMENTS; ACOUSTICS
- G10L—SPEECH ANALYSIS TECHNIQUES OR SPEECH SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING TECHNIQUES; SPEECH OR AUDIO CODING OR DECODING
- G10L15/00—Speech recognition
- G10L15/04—Segmentation; Word boundary detection
-
- G—PHYSICS
- G10—MUSICAL INSTRUMENTS; ACOUSTICS
- G10L—SPEECH ANALYSIS TECHNIQUES OR SPEECH SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING TECHNIQUES; SPEECH OR AUDIO CODING OR DECODING
- G10L15/00—Speech recognition
- G10L15/08—Speech classification or search
- G10L15/16—Speech classification or search using artificial neural networks
-
- G—PHYSICS
- G10—MUSICAL INSTRUMENTS; ACOUSTICS
- G10L—SPEECH ANALYSIS TECHNIQUES OR SPEECH SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING TECHNIQUES; SPEECH OR AUDIO CODING OR DECODING
- G10L15/00—Speech recognition
- G10L15/08—Speech classification or search
- G10L15/18—Speech classification or search using natural language modelling
- G10L15/1822—Parsing for meaning understanding
-
- G—PHYSICS
- G10—MUSICAL INSTRUMENTS; ACOUSTICS
- G10L—SPEECH ANALYSIS TECHNIQUES OR SPEECH SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING TECHNIQUES; SPEECH OR AUDIO CODING OR DECODING
- G10L15/00—Speech recognition
- G10L15/08—Speech classification or search
- G10L15/18—Speech classification or search using natural language modelling
- G10L15/183—Speech classification or search using natural language modelling using context dependencies, e.g. language models
-
- G—PHYSICS
- G10—MUSICAL INSTRUMENTS; ACOUSTICS
- G10L—SPEECH ANALYSIS TECHNIQUES OR SPEECH SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING TECHNIQUES; SPEECH OR AUDIO CODING OR DECODING
- G10L21/00—Speech or voice signal processing techniques to produce another audible or non-audible signal, e.g. visual or tactile, in order to modify its quality or its intelligibility
- G10L21/02—Speech enhancement, e.g. noise reduction or echo cancellation
-
- G—PHYSICS
- G10—MUSICAL INSTRUMENTS; ACOUSTICS
- G10L—SPEECH ANALYSIS TECHNIQUES OR SPEECH SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING TECHNIQUES; SPEECH OR AUDIO CODING OR DECODING
- G10L15/00—Speech recognition
- G10L15/22—Procedures used during a speech recognition process, e.g. man-machine dialogue
- G10L2015/226—Procedures used during a speech recognition process, e.g. man-machine dialogue using non-speech characteristics
Definitions
- the present invention relates to an artificial intelligence apparatus and a method for recognizing speech of a user in consideration of a user's application use. Specifically, the present invention relates to an artificial intelligence apparatus and a method for recognizing speech of a user more accurately by additionally considering a user's application use in a situation where the speech of the user is not correctly recognized.
- a speech recognition technology has been applied to various services such as a speech-to-text (STT) service that simply converts user's speech into text, a service that performs appropriate control or provides a response when the user's speech is input, or the like.
- STT speech-to-text
- the speech that has failed to be recognized once is often not recognized correctly even after the reutterance.
- the user will not use the speech recognition function without explicit feedback, which makes it difficult to improve a performance of the speech recognition function.
- a purpose of the present invention is to provide an artificial intelligence (AI) apparatus and a method for improving a speech recognition performance using an application usage log of a user as feedback when speech of the user is not correctly recognized.
- AI artificial intelligence
- Another purpose of the present invention to provide an AI apparatus and a method for updating a language model used for speech recognition in consideration of reliability of intention determination.
- an AI apparatus and a method for determining an intention of recognized speech of a user using an intention classifier and calculating reliability of the intention determination based on a relationship with locations of intention groups in a vector space projected by the intention classifier to determine whether to update a language model are provided.
- the speech recognition performance may be improved using the user's application usage log as an implicit feedback.
- user's specific pronunciation, user's specific language usage habit, dialect, or the like may be reflected to improve the speech recognition model.
- FIG. 1 is a block diagram illustrating an AI apparatus according to an embodiment of the present invention.
- FIG. 2 is a block diagram illustrating an AI server according to an embodiment of the present invention.
- FIG. 3 is a diagram illustrating an AI system according to an embodiment of the present invention.
- FIG. 4 is a block diagram illustrating an AI apparatus according to an embodiment of the present invention.
- FIG. 5 is a flowchart illustrating a method for recognizing speech of a user in consideration of user's application usage according to an embodiment of the present invention.
- FIG. 6 is a diagram illustrating a process of determining whether user's intention recognition is successful, according to an embodiment of the present invention.
- FIG. 7 is a flowchart illustrating an example of updating of a language model (S 513 ) shown in FIG. 5
- FIGS. 8 and 9 are diagrams illustrating a method for recognizing speech of a user according to an embodiment of the present invention.
- Machine learning refers to the field of defining various issues dealt with in the field of artificial intelligence and studying methodology for solving the various issues.
- Machine learning is defined as an algorithm that enhances the performance of a certain task through a steady experience with the certain task.
- An artificial neural network is a model used in machine learning and may mean a whole model of problem-solving ability which is composed of artificial neurons (nodes) that form a network by synaptic connections.
- the artificial neural network can be defined by a connection pattern between neurons in different layers, a learning process for updating model parameters, and an activation function for generating an output value.
- the artificial neural network may include an input layer, an output layer, and optionally one or more hidden layers. Each layer includes one or more neurons, and the artificial neural network may include a synapse that links neurons to neurons. In the artificial neural network, each neuron may output the function value of the activation function for input signals, weights, and deflections input through the synapse.
- Model parameters refer to parameters determined through learning and include a weight value of synaptic connection and deflection of neurons.
- a hyperparameter means a parameter to be set in the machine learning algorithm before learning, and includes a learning rate, a repetition number, a mini batch size, and an initialization function.
- the purpose of the learning of the artificial neural network may be to determine the model parameters that minimize a loss function.
- the loss function may be used as an index to determine optimal model parameters in the learning process of the artificial neural network.
- Machine learning may be classified into supervised learning, unsupervised learning, and reinforcement learning according to a learning method.
- the supervised learning may refer to a method of learning an artificial neural network in a state in which a label for training data is given, and the label may mean the correct answer (or result value) that the artificial neural network must infer when the training data is input to the artificial neural network.
- the unsupervised learning may refer to a method of learning an artificial neural network in a state in which a label for training data is not given.
- the reinforcement learning may refer to a learning method in which an agent defined in a certain environment learns to select a behavior or a behavior sequence that maximizes cumulative compensation in each state.
- Machine learning which is implemented as a deep neural network (DNN) including a plurality of hidden layers among artificial neural networks, is also referred to as deep learning, and the deep learning is part of machine learning.
- DNN deep neural network
- machine learning is used to mean deep learning.
- a robot may refer to a machine that automatically processes or operates a given task by its own ability.
- a robot having a function of recognizing an environment and performing a self-determination operation may be referred to as an intelligent robot.
- Robots may be classified into industrial robots, medical robots, home robots, military robots, and the like according to the use purpose or field.
- the robot includes a driving unit may include an actuator or a motor and may perform various physical operations such as moving a robot joint.
- a movable robot may include a wheel, a brake, a propeller, and the like in a driving unit, and may travel on the ground through the driving unit or fly in the air.
- Self-driving refers to a technique of driving for oneself, and a self-driving vehicle refers to a vehicle that travels without an operation of a user or with a minimum operation of a user.
- the self-driving may include a technology for maintaining a lane while driving, a technology for automatically adjusting a speed, such as adaptive cruise control, a technique for automatically traveling along a predetermined route, and a technology for automatically setting and traveling a route when a destination is set.
- the vehicle may include a vehicle having only an internal combustion engine, a hybrid vehicle having an internal combustion engine and an electric motor together, and an electric vehicle having only an electric motor, and may include not only an automobile but also a train, a motorcycle, and the like.
- the self-driving vehicle may be regarded as a robot having a self-driving function.
- Extended reality is collectively referred to as virtual reality (VR), augmented reality (AR), and mixed reality (MR).
- VR virtual reality
- AR augmented reality
- MR mixed reality
- the VR technology provides a real-world object and background only as a CG image
- the AR technology provides a virtual CG image on a real object image
- the MR technology is a computer graphic technology that mixes and combines virtual objects into the real world.
- the MR technology is similar to the AR technology in that the real object and the virtual object are shown together.
- the virtual object is used in the form that complements the real object, whereas in the MR technology, the virtual object and the real object are used in an equal manner.
- the XR technology may be applied to a head-mount display (HMD), a head-up display (HUD), a mobile phone, a tablet PC, a laptop, a desktop, a TV, a digital signage, and the like.
- HMD head-mount display
- HUD head-up display
- a device to which the XR technology is applied may be referred to as an XR device.
- FIG. 1 is a block diagram illustrating an AI apparatus 100 according to an embodiment of the present invention.
- the AI apparatus (or an AI device) 100 may be implemented by a stationary device or a mobile device, such as a TV, a projector, a mobile phone, a smartphone, a desktop computer, a notebook, a digital broadcasting terminal, a personal digital assistant (PDA), a portable multimedia player (PMP), a navigation device, a tablet PC, a wearable device, a set-top box (STB), a DMB receiver, a radio, a washing machine, a refrigerator, a desktop computer, a digital signage, a robot, a vehicle, and the like.
- a stationary device or a mobile device such as a TV, a projector, a mobile phone, a smartphone, a desktop computer, a notebook, a digital broadcasting terminal, a personal digital assistant (PDA), a portable multimedia player (PMP), a navigation device, a tablet PC, a wearable device, a set-top box (STB), a DMB receiver, a radio, a washing machine, a refrigerator
- the AI apparatus 100 may include a communication unit 110 , an input unit 120 , a learning processor 130 , a sensing unit 140 , an output unit 150 , a memory 170 , and a processor 180 .
- the communication unit 110 may transmit and receive data to and from external devices such as other AI apparatuses 100 a to 100 e and the AI server 200 by using wire/wireless communication technology.
- the communication unit 110 may transmit and receive sensor information, a user input, a learning model, and a control signal to and from external devices.
- the communication technology used by the communication unit 110 includes GSM (Global System for Mobile communication), CDMA (Code Division Multi Access), LTE (Long Term Evolution), 5G, WLAN (Wireless LAN), Wi-Fi (Wireless-Fidelity), BluetoothTM, RFID (Radio Frequency Identification), Infrared Data Association (IrDA), ZigBee, NFC (Near Field Communication), and the like.
- GSM Global System for Mobile communication
- CDMA Code Division Multi Access
- LTE Long Term Evolution
- 5G Fifth Generation
- WLAN Wireless LAN
- Wi-Fi Wireless-Fidelity
- BluetoothTM BluetoothTM
- RFID Radio Frequency Identification
- IrDA Infrared Data Association
- ZigBee ZigBee
- NFC Near Field Communication
- the input unit 120 may acquire various kinds of data.
- the input unit 120 may include a camera for inputting a video signal, a microphone for receiving an audio signal, and a user input unit for receiving information from a user.
- the camera or the microphone may be treated as a sensor, and the signal acquired from the camera or the microphone may be referred to as sensing data or sensor information.
- the input unit 120 may acquire a training data for model learning and an input data to be used when an output is acquired by using learning model.
- the input unit 120 may acquire raw input data.
- the processor 180 or the learning processor 130 may extract an input feature by preprocessing the input data.
- the learning processor 130 may learn a model composed of an artificial neural network by using training data.
- the learned artificial neural network may be referred to as a learning model.
- the learning model may be used to an infer result value for new input data rather than training data, and the inferred value may be used as a basis for determination to perform a certain operation.
- the learning processor 130 may perform AI processing together with the learning processor 240 of the AI server 200 .
- the learning processor 130 may include a memory integrated or implemented in the AI apparatus 100 .
- the learning processor 130 may be implemented by using the memory 170 , an external memory directly connected to the AI apparatus 100 , or a memory held in an external device.
- the sensing unit 140 may acquire at least one of internal information about the AI apparatus 100 , ambient environment information about the AI apparatus 100 , or user information by using various sensors.
- Examples of the sensors included in the sensing unit 140 may include a proximity sensor, an illuminance sensor, an acceleration sensor, a magnetic sensor, a gyro sensor, an inertial sensor, an RGB sensor, an IR sensor, a fingerprint recognition sensor, an ultrasonic sensor, an optical sensor, a microphone, a lidar, and a radar.
- a proximity sensor an illuminance sensor, an acceleration sensor, a magnetic sensor, a gyro sensor, an inertial sensor, an RGB sensor, an IR sensor, a fingerprint recognition sensor, an ultrasonic sensor, an optical sensor, a microphone, a lidar, and a radar.
- the output unit 150 may generate an output related to a visual sense, an auditory sense, or a haptic sense.
- the output unit 150 may include a display unit for outputting time information, a speaker for outputting auditory information, and a haptic module for outputting haptic information.
- the memory 170 may store data that supports various functions of the AI apparatus 100 .
- the memory 170 may store input data acquired by the input unit 120 , training data, a learning model, a learning history, and the like.
- the processor 180 may determine at least one executable operation of the AI apparatus 100 based on information determined or generated by using a data analysis algorithm or a machine learning algorithm.
- the processor 180 may control the components of the AI apparatus 100 to execute the determined operation.
- the processor 180 may request, search, receive, or utilize data of the learning processor 130 or the memory 170 .
- the processor 180 may control the components of the AI apparatus 100 to execute the predicted operation or the operation determined to be desirable among the at least one executable operation.
- the processor 180 may generate a control signal for controlling the external device and may transmit the generated control signal to the external device.
- the processor 180 may acquire intention information for the user input and may determine the user's requirements based on the acquired intention information.
- the processor 180 may acquire the intention information corresponding to the user input by using at least one of a speech to text (STT) engine for converting speech input into a text string or a natural language processing (NLP) engine for acquiring intention information of a natural language.
- STT speech to text
- NLP natural language processing
- At least one of the STT engine or the NLP engine may be configured as an artificial neural network, at least part of which is learned according to the machine learning algorithm. At least one of the STT engine or the NLP engine may be learned by the learning processor 130 , may be learned by the learning processor 240 of the AI server 200 , or may be learned by their distributed processing.
- the processor 180 may collect history information including the operation contents of the AI apparatus 100 or the user's feedback on the operation and may store the collected history information in the memory 170 or the learning processor 130 or transmit the collected history information to the external device such as the AI server 200 .
- the collected history information may be used to update the learning model.
- the processor 180 may control at least part of the components of AI apparatus 100 so as to drive an application program stored in memory 170 . Furthermore, the processor 180 may operate two or more of the components included in the AI apparatus 100 in combination so as to drive the application program.
- FIG. 2 is a block diagram illustrating an AI server 200 according to an embodiment of the present invention.
- the AI server 200 may refer to a device that learns an artificial neural network by using a machine learning algorithm or uses a learned artificial neural network.
- the AI server 200 may include a plurality of servers to perform distributed processing, or may be defined as a 5G network.
- the AI server 200 may be included as a partial configuration of the AI apparatus 100 , and may perform at least part of the AI processing together.
- the AI server 200 may include a communication unit 210 , a memory 230 , a learning processor 240 , a processor 260 , and the like.
- the communication unit 210 can transmit and receive data to and from an external device such as the AI apparatus 100 .
- the memory 230 may include a model storage unit 231 .
- the model storage unit 231 may store a learning or learned model (or an artificial neural network 231 a ) through the learning processor 240 .
- the learning processor 240 may learn the artificial neural network 231 a by using the training data.
- the learning model may be used in a state of being mounted on the AI server 200 of the artificial neural network, or may be used in a state of being mounted on an external device such as the AI apparatus 100 .
- the learning model may be implemented in hardware, software, or a combination of hardware and software. If all or part of the learning models are implemented in software, one or more instructions that constitute the learning model may be stored in memory 230 .
- the processor 260 may infer the result value for new input data by using the learning model and may generate a response or a control command based on the inferred result value.
- FIG. 3 is a diagram illustrating an AI system 1 according to an embodiment of the present invention.
- an AI server 200 a robot 100 a, a self-driving vehicle 100 b, an XR device 100 c, a smartphone 100 d, or a home appliance 100 e is connected to a cloud network 10 .
- the robot 100 a, the self-driving vehicle 100 b, the XR device 100 c, the smartphone 100 d, or the home appliance 100 e, to which the AI technology is applied, may be referred to as AI apparatuses 100 a to 100 e.
- the cloud network 10 may refer to a network that forms part of a cloud computing infrastructure or exists in a cloud computing infrastructure.
- the cloud network 10 may be configured by using a 3G network, a 4G or LTE network, or a 5G network.
- the devices 100 a to 100 e and 200 configuring the AI system 1 may be connected to each other through the cloud network 10 .
- each of the devices 100 a to 100 e and 200 may communicate with each other through a base station, but may directly communicate with each other without using a base station.
- the AI server 200 may include a server that performs AI processing and a server that performs operations on big data.
- the AI server 200 may be connected to at least one of the AI apparatuses constituting the AI system 1 , that is, the robot 100 a, the self-driving vehicle 100 b, the XR device 100 c, the smartphone 100 d, or the home appliance 100 e through the cloud network 10 , and may assist at least part of AI processing of the connected AI apparatuses 100 a to 100 e.
- the AI server 200 may learn the artificial neural network according to the machine learning algorithm instead of the AI apparatuses 100 a to 100 e, and may directly store the learning model or transmit the learning model to the AI apparatuses 100 a to 100 e.
- the AI server 200 may receive input data from the AI apparatuses 100 a to 100 e, may infer the result value for the received input data by using the learning model, may generate a response or a control command based on the inferred result value, and may transmit the response or the control command to the AI apparatuses 100 a to 100 e.
- the AI apparatuses 100 a to 100 e may infer the result value for the input data by directly using the learning model, and may generate the response or the control command based on the inference result.
- the AI apparatuses 100 a to 100 e illustrated in FIG. 3 may be regarded as a specific embodiment of the AI apparatus 100 illustrated in FIG. 1 .
- the robot 100 a may be implemented as a guide robot, a carrying robot, a cleaning robot, a wearable robot, an entertainment robot, a pet robot, an unmanned flying robot, or the like.
- the robot 100 a may include a robot control module for controlling the operation, and the robot control module may refer to a software module or a chip implementing the software module by hardware.
- the robot 100 a may acquire state information about the robot 100 a by using sensor information acquired from various kinds of sensors, may detect (recognize) surrounding environment and objects, may generate map data, may determine the route and the travel plan, may determine the response to user interaction, or may determine the operation.
- the robot 100 a may use the sensor information acquired from at least one sensor among the lidar, the radar, and the camera so as to determine the travel route and the travel plan.
- the robot 100 a may perform the above-described operations by using the learning model composed of at least one artificial neural network.
- the robot 100 a may recognize the surrounding environment and the objects by using the learning model, and may determine the operation by using the recognized surrounding information or object information.
- the learning model may be learned directly from the robot 100 a or may be learned from an external device such as the AI server 200 .
- the robot 100 a may perform the operation by generating the result by directly using the learning model, but the sensor information may be transmitted to the external device such as the AI server 200 and the generated result may be received to perform the operation.
- the robot 100 a may use at least one of the map data, the object information detected from the sensor information, or the object information acquired from the external device to determine the travel route and the travel plan, and may control the driving unit such that the robot 100 a travels along the determined travel route and travel plan.
- the map data may include object identification information about various objects arranged in the space in which the robot 100 a moves.
- the map data may include object identification information about fixed objects such as walls and doors and movable objects such as pollen and desks.
- the object identification information may include a name, a type, a distance, and a position.
- the robot 100 a may perform the operation or travel by controlling the driving unit based on the control/interaction of the user. At this time, the robot 100 a may acquire the intention information of the interaction due to the user's operation or speech utterance, and may determine the response based on the acquired intention information, and may perform the operation.
- the self-driving vehicle 100 b to which the AI technology is applied, may be implemented as a mobile robot, a vehicle, an unmanned flying vehicle, or the like.
- the self-driving vehicle 100 b may include a self-driving control module for controlling a self-driving function, and the self-driving control module may refer to a software module or a chip implementing the software module by hardware.
- the self-driving control module may be included in the self-driving vehicle 100 b as a component thereof, but may be implemented with separate hardware and connected to the outside of the self-driving vehicle 100 b.
- the self-driving vehicle 100 b may acquire state information about the self-driving vehicle 100 b by using sensor information acquired from various kinds of sensors, may detect (recognize) surrounding environment and objects, may generate map data, may determine the route and the travel plan, or may determine the operation.
- the self-driving vehicle 100 b may use the sensor information acquired from at least one sensor among the lidar, the radar, or the camera so as to determine the travel route and the travel plan.
- the self-driving vehicle 100 b may recognize the environment or objects for an area covered by a field of view or an area over a certain distance by receiving the sensor information from external devices, or may receive directly recognized information from the external devices.
- the self-driving vehicle 100 b may perform the above-described operations by using the learning model composed of at least one artificial neural network.
- the self-driving vehicle 100 b may recognize the surrounding environment and the objects by using the learning model, and may determine the traveling movement line by using the recognized surrounding information or object information.
- the learning model may be learned directly from the self-driving vehicle 100 a or may be learned from an external device such as the AI server 200 .
- the self-driving vehicle 100 b may perform the operation by generating the result by directly using the learning model, but the sensor information may be transmitted to the external device such as the AI server 200 and the generated result may be received to perform the operation.
- the self-driving vehicle 100 b may use at least one of the map data, the object information detected from the sensor information, or the object information acquired from the external device to determine the travel route and the travel plan, and may control the driving unit such that the self-driving vehicle 100 b travels along the determined travel route and travel plan.
- the map data may include object identification information about various objects arranged in the space (for example, road) in which the self-driving vehicle 100 b travels.
- the map data may include object identification information about fixed objects such as street lamps, rocks, and buildings and movable objects such as vehicles and pedestrians.
- the object identification information may include a name, a type, a distance, and a position.
- the self-driving vehicle 100 b may perform the operation or travel by controlling the driving unit based on the control/interaction of the user. At this time, the self-driving vehicle 100 b may acquire the intention information of the interaction due to the user's operation or speech utterance, and may determine the response based on the acquired intention information, and may perform the operation.
- the XR device 100 c may be implemented by a head-mount display (HMD), a head-up display (HUD) provided in the vehicle, a television, a mobile phone, a smartphone, a computer, a wearable device, a home appliance, a digital signage, a vehicle, a fixed robot, a mobile robot, or the like.
- HMD head-mount display
- HUD head-up display
- the XR device 100 c may analyzes three-dimensional point cloud data or image data acquired from various sensors or the external devices, generate position data and attribute data for the three-dimensional points, acquire information about the surrounding space or the real object, and render to output the XR object to be output. For example, the XR device 100 c may output an XR object including the additional information about the recognized object in correspondence to the recognized object.
- the XR device 100 c may perform the above-described operations by using the learning model composed of at least one artificial neural network. For example, the XR device 100 c may recognize the real object from the three-dimensional point cloud data or the image data by using the learning model, and may provide information corresponding to the recognized real object.
- the learning model may be directly learned from the XR device 100 c, or may be learned from the external device such as the AI server 200 .
- the XR device 100 c may perform the operation by generating the result by directly using the learning model, but the sensor information may be transmitted to the external device such as the AI server 200 and the generated result may be received to perform the operation.
- the robot 100 a may be implemented as a guide robot, a carrying robot, a cleaning robot, a wearable robot, an entertainment robot, a pet robot, an unmanned flying robot, or the like.
- the robot 100 a to which the AI technology and the self-driving technology are applied, may refer to the robot itself having the self-driving function or the robot 100 a interacting with the self-driving vehicle 100 b.
- the robot 100 a having the self-driving function may collectively refer to a device that moves for itself along the given movement line without the user's control or moves for itself by determining the movement line by itself.
- the robot 100 a and the self-driving vehicle 100 b having the self-driving function may use a common sensing method so as to determine at least one of the travel route or the travel plan.
- the robot 100 a and the self-driving vehicle 100 b having the self-driving function may determine at least one of the travel route or the travel plan by using the information sensed through the lidar, the radar, and the camera.
- the robot 100 a that interacts with the self-driving vehicle 100 b exists separately from the self-driving vehicle 100 b and may perform operations interworking with the self-driving function of the self-driving vehicle 100 b or interworking with the user who rides on the self-driving vehicle 100 b.
- the robot 100 a interacting with the self-driving vehicle 100 b may control or assist the self-driving function of the self-driving vehicle 100 b by acquiring sensor information on behalf of the self-driving vehicle 100 b and providing the sensor information to the self-driving vehicle 100 b, or by acquiring sensor information, generating environment information or object information, and providing the information to the self-driving vehicle 100 b.
- the robot 100 a interacting with the self-driving vehicle 100 b may monitor the user boarding the self-driving vehicle 100 b, or may control the function of the self-driving vehicle 100 b through the interaction with the user. For example, when it is determined that the driver is in a drowsy state, the robot 100 a may activate the self-driving function of the self-driving vehicle 100 b or assist the control of the driving unit of the self-driving vehicle 100 b.
- the function of the self-driving vehicle 100 b controlled by the robot 100 a may include not only the self-driving function but also the function provided by the navigation system or the audio system provided in the self-driving vehicle 100 b.
- the robot 100 a that interacts with the self-driving vehicle 100 b may provide information or assist the function to the self-driving vehicle 100 b outside the self-driving vehicle 100 b.
- the robot 100 a may provide traffic information including signal information and the like, such as a smart signal, to the self-driving vehicle 100 b, and automatically connect an electric charger to a charging port by interacting with the self-driving vehicle 100 b like an automatic electric charger of an electric vehicle.
- the robot 100 a may be implemented as a guide robot, a carrying robot, a cleaning robot, a wearable robot, an entertainment robot, a pet robot, an unmanned flying robot, a drone, or the like.
- the robot 100 a to which the XR technology is applied, may refer to a robot that is subjected to control/interaction in an XR image.
- the robot 100 a may be separated from the XR device 100 c and interwork with each other.
- the robot 100 a When the robot 100 a, which is subjected to control/interaction in the XR image, may acquire the sensor information from the sensors including the camera, the robot 100 a or the XR device 100 c may generate the XR image based on the sensor information, and the XR device 100 c may output the generated XR image.
- the robot 100 a may operate based on the control signal input through the XR device 100 c or the user's interaction.
- the user can confirm the XR image corresponding to the time point of the robot 100 a interworking remotely through the external device such as the XR device 100 c, adjust the self-driving travel path of the robot 100 a through interaction, control the operation or driving, or confirm the information about the surrounding object.
- the external device such as the XR device 100 c
- the self-driving vehicle 100 b to which the AI technology and the XR technology are applied, may be implemented as a mobile robot, a vehicle, an unmanned flying vehicle, or the like.
- the self-driving driving vehicle 100 b may refer to a self-driving vehicle having a means for providing an XR image or a self-driving vehicle that is subjected to control/interaction in an XR image.
- the self-driving vehicle 100 b that is subjected to control/interaction in the XR image may be distinguished from the XR device 100 c and interwork with each other.
- the self-driving vehicle 100 b having the means for providing the XR image may acquire the sensor information from the sensors including the camera and output the generated XR image based on the acquired sensor information.
- the self-driving vehicle 100 b may include an HUD to output an XR image, thereby providing a passenger with a real object or an XR object corresponding to an object in the screen.
- the self-driving vehicle 100 b may output XR objects corresponding to objects such as a lane, another vehicle, a traffic light, a traffic sign, a two-wheeled vehicle, a pedestrian, a building, and the like.
- the self-driving vehicle 100 b When the self-driving vehicle 100 b, which is subjected to control/interaction in the XR image, may acquire the sensor information from the sensors including the camera, the self-driving vehicle 100 b or the XR device 100 c may generate the XR image based on the sensor information, and the XR device 100 c may output the generated XR image.
- the self-driving vehicle 100 b may operate based on the control signal input through the external device such as the XR device 100 c or the user's interaction.
- FIG. 4 is a block diagram illustrating an AI apparatus according to an embodiment of the present invention.
- the input unit 120 may include a camera 121 for image signal input, a microphone 122 for receiving audio signal input, and a user input unit 123 for receiving information from a user.
- Voice data or image data collected by the input unit 120 are analyzed and processed as a user's control command.
- the input unit 120 is used for inputting image information (or signal), audio information (or signal), data, or information inputted from a user and the AI apparatus 100 may include at least one camera 121 in order for inputting image information.
- the camera 121 processes image frames such as a still image or a video obtained by an image sensor in a video call mode or a capturing mode.
- the processed image frame may be displayed on the display unit 151 or stored in the memory 170 .
- the microphone 122 processes external sound signals as electrical voice data.
- the processed voice data may be utilized variously according to a function (or an application program being executed) being performed in the AI apparatus 100 .
- various noise canceling algorithms for removing noise occurring during the reception of external sound signals may be implemented in the microphone 122 .
- the user input unit 123 is to receive information from a user and when information is inputted through the user input unit 123 , the processor 180 may control an operation of the AI apparatus 100 to correspond to the inputted information.
- the user input unit 123 may include a mechanical input means (or a mechanical key, for example, a button, a dome switch, a jog wheel, and a jog switch at the front, back or side of the AI apparatus 100 ) and a touch type input means.
- a touch type input means may include a virtual key, a soft key, or a visual key, which is displayed on a touch screen through software processing or may include a touch key disposed at a portion other than the touch screen.
- a sensing unit 140 may be called a sensor unit.
- the output unit 150 may include at least one of a display unit 151 , a sound output module 152 , a haptic module 153 , or an optical output module 154 .
- the display unit 151 may display (output) information processed in the AI apparatus 100 .
- the display unit 151 may display execution screen information of an application program running on the AI apparatus 100 or user interface (UI) and graphic user interface (GUI) information according to such execution screen information.
- UI user interface
- GUI graphic user interface
- the display unit 151 may be formed with a mutual layer structure with a touch sensor or formed integrally, so that a touch screen may be implemented.
- a touch screen may serve as the user input unit 123 providing an input interface between the AI apparatus 100 and a user, and an output interface between the AI apparatus 100 and a user at the same time.
- the sound output module 152 may output audio data received from the wireless communication unit 110 or stored in the memory 170 in a call signal reception or call mode, a recording mode, a voice recognition mode, or a broadcast reception mode.
- the sound output module 152 may include a receiver, a speaker, and a buzzer.
- the haptic module 153 generates various haptic effects that a user can feel.
- a representative example of a haptic effect that the haptic module 153 generates is vibration.
- the optical output module 154 outputs a signal for notifying event occurrence by using light of a light source of the AI apparatus 100 .
- An example of an event occurring in the AI apparatus 100 includes message reception, call signal reception, missed calls, alarm, schedule notification, e-mail reception, and information reception through an application.
- FIG. 5 is a flowchart illustrating a method for recognizing speech of a user in consideration of user's application usage according to an embodiment of the present invention.
- the processor 180 of the AI apparatus 100 receives a sound signal for the speech of the user (S 501 ).
- the sound signal of the user's speech may be received via the microphone 121 of the AI apparatus 100 , but may also be received from an external device (not shown) via the communication unit 110 .
- the sound signal may be an audio file in a pulse code modulation (PCM) format.
- PCM pulse code modulation
- the processor 180 of the AI apparatus 100 recognizes the speech using a language model (LM) (S 503 ).
- LLM language model
- the recognizing of the speech may mean generating a text string corresponding to the sound signal.
- the processor 180 may recognize the input sound signal on a phoneme basis using an acoustic model (AM) and may determine a word corresponding to the input sound signal based on the recognition result on a phoneme basis.
- the processor 180 may generate a word vector on a phoneme basis using the recognition result on a phoneme basis and determine at least one word corresponding to the input sound signal using the generated word vector on a phoneme basis and the language model.
- AM acoustic model
- the at least one word determined by the processor 180 may include a single word or a word sequence composed of a plurality of words.
- the language model may refer to a model that represents a probability distribution for the word sequence. That is, when a specific word sequence is given, the language model may output a probability that the word sequence appears. Thus, the processor 180 may use the language model to determine candidate words suitable for a next word when a specific word sequence is given or to calculate a probability that each candidate word appears.
- the processor 180 recognizes phonemes of a specific recognition target word as [‘n’, ‘i’, ‘g’, ‘h’, ‘t’] using the acoustic model.
- the processor 180 may generate a word vector on a phoneme basis corresponding to the [‘n’, ‘i’, ‘g’, ‘h’, ‘t’] and determine the recognition target word using the generated word vector and the language model.
- the processor 180 may determine the target word to be the “knight” rather than the “night”, unlike the result of the acoustic model.
- the processor 180 may determine a single word using word vectors on a phoneme basis estimated to be a plurality of words, or conversely, determine a plurality of words using a word vector on a phoneme basis estimated to be a single word.
- the recognition target word may be determined as a plurality of words “for him”.
- the recognition target word may be determined as the single word “foreign” based on the determination result of the language model.
- the processor 180 may recognize a word based on the recognition result on a phoneme basis of the acoustic model and calculate an LM score of the language model for the recognized word. Further, when the LM score of the recognized word is below a certain level, the processor 180 may find a word similar to the recognized word but has a higher LM score from a lexicon and determine the found word to be a recognition result.
- the processor 180 may recognize the word more accurately using the word vector on a phoneme basis and the language model for the word, which is the current recognition target, together. Thus, the processor 180 may more accurately recognize the speech of the user in general.
- the language model may be learned as a personalized model classified for each user, and the processor 180 may recognize the user when the sound signal is received and recognize the speech using the personalized language model corresponding to the recognized user.
- Each of the acoustic model or the language model described above may be a model learned using a machine learning algorithm or a deep learning algorithm, and may be configured as an artificial neural network. Learning of the acoustic model or the language model may be performed in the learning processor 130 of the AI apparatus 100 or in the learning processor 240 of the AI server 200 .
- the acoustic model or the language model may be stored in the memory 170 of the AI apparatus 100 or may be stored in the memory 230 of the AI server 200 .
- the processor 180 of the AI apparatus 100 determines the intention of the user based on the speech recognition result (S 505 ).
- the determining of the intention of the user may mean recognizing the intention of the user. Further, the determining of the intention corresponding to the user's speech may mean determining the intention of the natural language corresponding to the speech. Further, the intention of the user may mean an intention corresponding to the user's speech.
- the processor 180 may determine the intention of the user by generating the intention information based on the speech recognition result.
- the processor 180 may determine the intention of the user using an intention classifier of a natural language understanding (NLU) technique.
- NLU natural language understanding
- the intention classifier may refer to a model which projects an intention of the input speech recognition result onto a vector space, when the speech recognition result is inputted.
- a cluster may be formed for each intention in the vector space resulted from an output of the intention classifier. That is, there may be a plurality of intention clusters that may be classified by the intention classifier in the vector space resulted from the output of the intention classifier.
- the intention classifier may project the input speech recognition result to a location adjacent to the corresponding intention cluster. Therefore, a distance between the projected location and each intention cluster means a similarity between the intention of the input speech recognition result and each intention cluster. Thus, a short distance may mean a high degree of the similarity.
- the intention classifier may be an artificial neural network-based model learned using the machine learning algorithm or the deep learning algorithm.
- the processor 180 of the AI apparatus 100 determines whether the intention recognition is successful (S 507 ).
- Whether the intention recognition is successful may mean whether accuracy or reliability of the intention recognition exceeds a preset first reference value.
- the processor 180 may calculate the reliability of the intention recognition using the distance from the location to which the speech recognition result input in the vector space is projected by the intention classifier to each of the intention clusters.
- the processor 180 may calculate the reliability of the intention recognition using a distance from the location to which the speech recognition result input in the vector space is projected to the nearest intention cluster. This is because an intention corresponding to the intention cluster nearest to the location to which the input speech recognition result is projected may be determined to be the intention of the input speech recognition result.
- the distance from the location to which the input speech recognition result is projected to the nearest intention cluster may be referred to as a first cluster distance.
- a distance from the location to which the input speech recognition result is projected to an intention cluster distant at an n-th spacing may be referred to as an n-th cluster distance.
- the intention cluster distant at an n-th spacing from the location to which the input speech recognition result is projected may be referred to as an n-th cluster.
- the processor 180 may calculate higher reliability of the intention recognition. Further, the processor 180 may determine whether the intention recognition is successful based on whether the calculated reliability exceeds the first reference value.
- the processor 180 may calculate the reliability of the intention recognition using both the first cluster distance and the second cluster distance.
- the processor 180 may calculate higher reliability of the intention recognition. Further, the processor 180 may determine whether the intention recognition is successful based on whether the calculated reliability exceeds the first reference value. This is because it may be seen that the difference between the first cluster distance and the second cluster distance indicates how assuredly the intention analyzer determined the intention of the input speech recognition result to be an intention that corresponds to the first cluster.
- the processor 180 may determine whether the intention recognition is successful by determining whether the first cluster distance is smaller than a preset second reference value without explicitly calculating the reliability of the intention recognition.
- the processor 180 may determine whether the intention recognition is successful by determining whether the first cluster distance is smaller than a preset third reference value and whether a difference between the first cluster distance and the second distance is larger than a preset fourth reference value.
- the processor 180 of the AI apparatus 100 performs an operation corresponding to the determined intention of the user (S 509 ).
- the processor 180 may output corresponding information via the speaker or the display unit.
- the processor 180 of the AI apparatus 100 obtains a user's application usage log (S 511 ).
- the application usage log may include not only an application usage log in the AI apparatus 100 but also an application usage log in an external device capable of communicating with the AI apparatus 100 .
- the application usage log may include a type of an executed application, a content of user's manipulation for the executed application, and the like.
- the content of the user's manipulation for the executed application may include a search history of the user, a browse history of the search result of the user, and the like.
- the processor 180 may obtain the application usage log before and after the user's utterance. This is because a pattern of the usage of the application before and after the user's utterance may be expected to be highly related to utterance content.
- the user's question is likely to be a question about the weather.
- the AI apparatus 100 did not grasp the intention of the user, and when the recognized word is similar to a movie title displayed in the movie ticket booking application, the user's question is likely to be a question asking for information about the corresponding movie.
- the application usage log before and after the user's utterance helps to grasp the content of the user's utterance and the intention of the utterance, and may be used as implicit feedback.
- the processor 180 may obtain application usage log for a predetermined period from an utterance time point.
- the predetermined period may be determined to be a period including the utterance time point, such as a time period from 30 seconds before the utterance time point to 1 minute after the utterance time or may be determined to be a period from the utterance time point.
- the processor 180 may determine whether an permission to obtain the application usage log exists. Further, when the obtaining permission does not exist, the processor 180 may output a notification requesting the obtaining permission.
- the processor 180 of the AI apparatus 100 updates the language model using the obtained user's application usage log (S 513 ).
- the processor 180 may set a word mapping relationship or adjust a weighted value, so that the language model corrects incorrectly recognized words based on the application usage log.
- the processor 180 recognized the speech of the user using the language model and the recognition result is “two dies weather” and it is assumed that the user runs the weather application or searches “today's weather” on the Internet.
- the processor 180 may map the “two dies” to a “today's” for the language model or increase a weighted value of the “today's”, so that the language model may recognize the “today's” rather than the “two dies” for the same speech later based on the application usage log.
- the processor 180 may update the language model by mapping incorrectly recognized words to popular words in consideration of the user's dialect, intonation, and unusual word habits.
- the setting of the mapping relationship of the words and the adjusting of the weighted value in the language model may mean word embedding learning.
- the processor 180 may compare the recognition result of the language model with the application usage log to determine whether the application usage log is related to the utterance of the user. Further, the language model may only be updated when it is determined that the application usage log is relevant to the utterance of the user.
- the processor 180 may extract a keyword to be used for the language model update from the application usage log based on the recognition result of the language model and the application usage log. Further, the processor 180 may find a word (or keyword) corresponding to the extracted keyword among the recognition results from the language model, map the corresponding word found to the extracted keyword for the language model, or increase a weighted value of the extracted keyword. In addition, the processor 180 may increase a weighted value of each of words having pronunciation similar to that of the extracted keyword.
- the processor 180 may update the language model for each user.
- the updated language model may be used to recognize future speech of the user.
- the language model may be used to recognize the user's speech, and may determine a word corresponding to the speech from the recognition result on a phoneme basis.
- the speech recognition performance of the general language model may deteriorate due to poor pronunciation, habit, or dialect of the user.
- a language model specific to an utterance feature may be constructed for each user by obtaining the implicit feedback for a situation where the speech recognition fails based on the user's application usage log and updating the language model for each user based on the implicit feedback.
- FIG. 6 is a diagram illustrating a process of determining whether user's intention recognition is successful, according to an embodiment of the present invention.
- the processor 180 of the AI apparatus 100 may determine whether the user's intention recognition is successful from a speech recognition result 610 using an intention classifier 620 of the natural language understanding technique.
- the intention classifier 620 may project the intention from the speech recognition result 610 input onto a vector space 630 . Further, a plurality of intention clusters that may be classified by the intention classifier 620 may exist in the vector space 630 .
- the processor 180 may calculate reliability of the intention recognition at a low level and thus determine that the recognition of the intention has failed.
- FIG. 7 is a flowchart illustrating an example of updating a language model (S 513 ) shown in FIG. 5 .
- the processor 180 of the AI apparatus 100 calculates correlation between the speech recognition result and the application usage log (S 701 ).
- the AI apparatus 100 does not correctly determine the intention corresponding to the user's speech, the user may show a pattern of the usage of the application, which is not related to the speech.
- the processor 180 may calculate the correlation between the user's application usage log and the speech recognition result to determine whether the application usage log may be used to update the language model.
- the correlation between the speech recognition result and the application usage log may mean a similarity between the speech recognition result and the application usage log.
- the processor 180 may determine a similarity of pronunciation and meaning between words or keywords included in the speech recognition result and words or keywords included in the application usage log, and may calculate the correlation between the speech recognition result and the application usage log based on the similarity of the pronunciation and meaning.
- the words or keywords included in the application usage log may include a keyword indicating a type and a name of the application included in the usage log, a keyword indicating information contained in the application, and the like.
- the processor 180 of the AI apparatus 100 determines whether the calculated correlation is equal to or above a preset reference value (S 703 ).
- the processor 180 of the AI apparatus 100 updates the language model (S 705 ).
- the processor 180 may determine a mapping relationship between the keyword of the speech recognition result and the keyword of the application usage log, and then the processor 180 may update the language model based on the determined mapping relationship.
- FIGS. 8 and 9 are diagrams illustrating a method for recognizing speech of a user according to an embodiment of the present invention.
- the user 811 may give speech utterance 812 “Today's weather” to the AI apparatus 821 to ask about today's weather.
- the processor of the AI apparatus 821 may incorrectly recognize the user's speech utterance 812 as “Two dies weather” ( 822 ).
- the processor may fail to grasp the intention of the user and suggest the user to search the web 823 .
- the “Today's” may be recognized as “Two dies” when an acoustic model and a language model based on English pronunciation of American speaker are used.
- the processor of the AI apparatus 821 may use the intention classifier to determine that the speech recognition result 822 in the vector space 831 is spaced apart from all the intention clusters. Thus, the processor may determine that the intention recognition has failed.
- the processor of the AI apparatus 821 may map 851 the “Two dies” to “Today's” using this application usage log, and update the language model based on this mapping relationship.
- the terminal 841 may be the same device as the AI apparatus 821 , but may be a separate device that is different from the AI apparatus 821 .
- FIG. 9 illustrates a process of recognizing speech of a user after a language model is updated according to FIG. 8 .
- the updated language model may recognize 922 the user's speech utterance 812 as “Today's weather” instead of “Two dies weather”. Accordingly, the AI apparatus 821 may provide the user with current weather information 923 .
- the above-described method may be implemented as a processor-readable code in a medium where a program is recorded.
- a processor-readable medium may include read-only memory (ROM), random access memory (RAM), CD-ROM, a magnetic tape, a floppy disk, and an optical data storage device.
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Health & Medical Sciences (AREA)
- Computational Linguistics (AREA)
- Audiology, Speech & Language Pathology (AREA)
- Human Computer Interaction (AREA)
- Acoustics & Sound (AREA)
- Multimedia (AREA)
- Artificial Intelligence (AREA)
- General Physics & Mathematics (AREA)
- Mathematical Physics (AREA)
- Theoretical Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Mathematical Analysis (AREA)
- Computational Mathematics (AREA)
- Mathematical Optimization (AREA)
- Pure & Applied Mathematics (AREA)
- Computing Systems (AREA)
- Evolutionary Computation (AREA)
- General Engineering & Computer Science (AREA)
- Software Systems (AREA)
- Databases & Information Systems (AREA)
- Quality & Reliability (AREA)
- Algebra (AREA)
- Signal Processing (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Life Sciences & Earth Sciences (AREA)
- Biomedical Technology (AREA)
- Biophysics (AREA)
- General Health & Medical Sciences (AREA)
- Molecular Biology (AREA)
- Navigation (AREA)
- User Interface Of Digital Computer (AREA)
Abstract
Description
- Pursuant to 35 U.S.C. § 119(a), this application claims the benefit of earlier filing date and right of priority to Korean Patent Application No. 10-2019-0088516, filed on Jul. 22, 2019, the contents of which are hereby incorporated by reference herein in its entirety.
- The present invention relates to an artificial intelligence apparatus and a method for recognizing speech of a user in consideration of a user's application use. Specifically, the present invention relates to an artificial intelligence apparatus and a method for recognizing speech of a user more accurately by additionally considering a user's application use in a situation where the speech of the user is not correctly recognized.
- Recently, the number of devices that interact with users using speech recognition are increasing. A speech recognition technology has been applied to various services such as a speech-to-text (STT) service that simply converts user's speech into text, a service that performs appropriate control or provides a response when the user's speech is input, or the like.
- However, when the user's speech is not recognized correctly, these devices will only ask the user to utter again, which causes a user's satisfaction level to drop significantly.
- In addition, the speech that has failed to be recognized once is often not recognized correctly even after the reutterance. In this case, the user will not use the speech recognition function without explicit feedback, which makes it difficult to improve a performance of the speech recognition function.
- A purpose of the present invention is to provide an artificial intelligence (AI) apparatus and a method for improving a speech recognition performance using an application usage log of a user as feedback when speech of the user is not correctly recognized.
- Further, another purpose of the present invention to provide an AI apparatus and a method for updating a language model used for speech recognition in consideration of reliability of intention determination.
- In a first aspect, there is provided an AI apparatus and a method for recognizing speech of a user using a language model, determining an intention of the user based on the recognition result, obtaining a user's application usage log when the determination of the intention is not successful, and updating the language model using the obtained application usage log.
- In a second aspect, there is provided an AI apparatus and a method for determining an intention of recognized speech of a user using an intention classifier and calculating reliability of the intention determination based on a relationship with locations of intention groups in a vector space projected by the intention classifier to determine whether to update a language model.
- According to various embodiments of the present invention, even when the speech of the user is not correctly recognized and the user does not provide explicit feedback, the speech recognition performance may be improved using the user's application usage log as an implicit feedback.
- Further, according to various embodiments of the present invention, user's specific pronunciation, user's specific language usage habit, dialect, or the like may be reflected to improve the speech recognition model.
-
FIG. 1 is a block diagram illustrating an AI apparatus according to an embodiment of the present invention. -
FIG. 2 is a block diagram illustrating an AI server according to an embodiment of the present invention. -
FIG. 3 is a diagram illustrating an AI system according to an embodiment of the present invention. -
FIG. 4 is a block diagram illustrating an AI apparatus according to an embodiment of the present invention. -
FIG. 5 is a flowchart illustrating a method for recognizing speech of a user in consideration of user's application usage according to an embodiment of the present invention. -
FIG. 6 is a diagram illustrating a process of determining whether user's intention recognition is successful, according to an embodiment of the present invention. -
FIG. 7 is a flowchart illustrating an example of updating of a language model (S513) shown inFIG. 5 -
FIGS. 8 and 9 are diagrams illustrating a method for recognizing speech of a user according to an embodiment of the present invention. - Hereinafter, embodiments of the present invention are described in more detail with reference to accompanying drawings and regardless of the drawings symbols, same or similar components are assigned with the same reference numerals and thus overlapping descriptions for those are omitted. The suffixes “module” and “unit” for components used in the description below are assigned or mixed in consideration of easiness in writing the specification and do not have distinctive meanings or roles by themselves. In the following description, detailed descriptions of well-known functions or constructions will be omitted since they would obscure the invention in unnecessary detail. Additionally, the accompanying drawings are used to help easily understanding embodiments disclosed herein but the technical idea of the present invention is not limited thereto. It should be understood that all of variations, equivalents or substitutes contained in the concept and technical scope of the present invention are also included.
- It will be understood that the terms “first” and “second” are used herein to describe various components but these components should not be limited by these terms. These terms are used only to distinguish one component from other components.
- In this disclosure below, when one part (or element, device, etc.) is referred to as being ‘connected’ to another part (or element, device, etc.), it should be understood that the former can be ‘directly connected’ to the latter, or ‘electrically connected’ to the latter via an intervening part (or element, device, etc.). It will be further understood that when one component is referred to as being ‘directly connected’ or ‘directly linked’ to another component, it means that no intervening component is present.
- <Artificial Intelligence (AI)>
- Artificial intelligence refers to the field of studying artificial intelligence or methodology for making artificial intelligence, and machine learning refers to the field of defining various issues dealt with in the field of artificial intelligence and studying methodology for solving the various issues. Machine learning is defined as an algorithm that enhances the performance of a certain task through a steady experience with the certain task.
- An artificial neural network (ANN) is a model used in machine learning and may mean a whole model of problem-solving ability which is composed of artificial neurons (nodes) that form a network by synaptic connections. The artificial neural network can be defined by a connection pattern between neurons in different layers, a learning process for updating model parameters, and an activation function for generating an output value.
- The artificial neural network may include an input layer, an output layer, and optionally one or more hidden layers. Each layer includes one or more neurons, and the artificial neural network may include a synapse that links neurons to neurons. In the artificial neural network, each neuron may output the function value of the activation function for input signals, weights, and deflections input through the synapse.
- Model parameters refer to parameters determined through learning and include a weight value of synaptic connection and deflection of neurons. A hyperparameter means a parameter to be set in the machine learning algorithm before learning, and includes a learning rate, a repetition number, a mini batch size, and an initialization function.
- The purpose of the learning of the artificial neural network may be to determine the model parameters that minimize a loss function. The loss function may be used as an index to determine optimal model parameters in the learning process of the artificial neural network.
- Machine learning may be classified into supervised learning, unsupervised learning, and reinforcement learning according to a learning method.
- The supervised learning may refer to a method of learning an artificial neural network in a state in which a label for training data is given, and the label may mean the correct answer (or result value) that the artificial neural network must infer when the training data is input to the artificial neural network. The unsupervised learning may refer to a method of learning an artificial neural network in a state in which a label for training data is not given. The reinforcement learning may refer to a learning method in which an agent defined in a certain environment learns to select a behavior or a behavior sequence that maximizes cumulative compensation in each state.
- Machine learning, which is implemented as a deep neural network (DNN) including a plurality of hidden layers among artificial neural networks, is also referred to as deep learning, and the deep learning is part of machine learning. In the following, machine learning is used to mean deep learning.
- <Robot>
- A robot may refer to a machine that automatically processes or operates a given task by its own ability. In particular, a robot having a function of recognizing an environment and performing a self-determination operation may be referred to as an intelligent robot.
- Robots may be classified into industrial robots, medical robots, home robots, military robots, and the like according to the use purpose or field.
- The robot includes a driving unit may include an actuator or a motor and may perform various physical operations such as moving a robot joint. In addition, a movable robot may include a wheel, a brake, a propeller, and the like in a driving unit, and may travel on the ground through the driving unit or fly in the air.
- <Self-Driving>
- Self-driving refers to a technique of driving for oneself, and a self-driving vehicle refers to a vehicle that travels without an operation of a user or with a minimum operation of a user.
- For example, the self-driving may include a technology for maintaining a lane while driving, a technology for automatically adjusting a speed, such as adaptive cruise control, a technique for automatically traveling along a predetermined route, and a technology for automatically setting and traveling a route when a destination is set.
- The vehicle may include a vehicle having only an internal combustion engine, a hybrid vehicle having an internal combustion engine and an electric motor together, and an electric vehicle having only an electric motor, and may include not only an automobile but also a train, a motorcycle, and the like.
- At this time, the self-driving vehicle may be regarded as a robot having a self-driving function.
- <eXtended Reality (XR)>
- Extended reality is collectively referred to as virtual reality (VR), augmented reality (AR), and mixed reality (MR). The VR technology provides a real-world object and background only as a CG image, the AR technology provides a virtual CG image on a real object image, and the MR technology is a computer graphic technology that mixes and combines virtual objects into the real world.
- The MR technology is similar to the AR technology in that the real object and the virtual object are shown together. However, in the AR technology, the virtual object is used in the form that complements the real object, whereas in the MR technology, the virtual object and the real object are used in an equal manner.
- The XR technology may be applied to a head-mount display (HMD), a head-up display (HUD), a mobile phone, a tablet PC, a laptop, a desktop, a TV, a digital signage, and the like. A device to which the XR technology is applied may be referred to as an XR device.
-
FIG. 1 is a block diagram illustrating anAI apparatus 100 according to an embodiment of the present invention. - The AI apparatus (or an AI device) 100 may be implemented by a stationary device or a mobile device, such as a TV, a projector, a mobile phone, a smartphone, a desktop computer, a notebook, a digital broadcasting terminal, a personal digital assistant (PDA), a portable multimedia player (PMP), a navigation device, a tablet PC, a wearable device, a set-top box (STB), a DMB receiver, a radio, a washing machine, a refrigerator, a desktop computer, a digital signage, a robot, a vehicle, and the like.
- Referring to
FIG. 1 , theAI apparatus 100 may include acommunication unit 110, aninput unit 120, a learningprocessor 130, asensing unit 140, anoutput unit 150, amemory 170, and aprocessor 180. - The
communication unit 110 may transmit and receive data to and from external devices such asother AI apparatuses 100 a to 100 e and theAI server 200 by using wire/wireless communication technology. For example, thecommunication unit 110 may transmit and receive sensor information, a user input, a learning model, and a control signal to and from external devices. - The communication technology used by the
communication unit 110 includes GSM (Global System for Mobile communication), CDMA (Code Division Multi Access), LTE (Long Term Evolution), 5G, WLAN (Wireless LAN), Wi-Fi (Wireless-Fidelity), Bluetooth™, RFID (Radio Frequency Identification), Infrared Data Association (IrDA), ZigBee, NFC (Near Field Communication), and the like. - The
input unit 120 may acquire various kinds of data. - At this time, the
input unit 120 may include a camera for inputting a video signal, a microphone for receiving an audio signal, and a user input unit for receiving information from a user. The camera or the microphone may be treated as a sensor, and the signal acquired from the camera or the microphone may be referred to as sensing data or sensor information. - The
input unit 120 may acquire a training data for model learning and an input data to be used when an output is acquired by using learning model. Theinput unit 120 may acquire raw input data. In this case, theprocessor 180 or thelearning processor 130 may extract an input feature by preprocessing the input data. - The learning
processor 130 may learn a model composed of an artificial neural network by using training data. The learned artificial neural network may be referred to as a learning model. The learning model may be used to an infer result value for new input data rather than training data, and the inferred value may be used as a basis for determination to perform a certain operation. - At this time, the learning
processor 130 may perform AI processing together with the learningprocessor 240 of theAI server 200. - At this time, the learning
processor 130 may include a memory integrated or implemented in theAI apparatus 100. Alternatively, the learningprocessor 130 may be implemented by using thememory 170, an external memory directly connected to theAI apparatus 100, or a memory held in an external device. - The
sensing unit 140 may acquire at least one of internal information about theAI apparatus 100, ambient environment information about theAI apparatus 100, or user information by using various sensors. - Examples of the sensors included in the
sensing unit 140 may include a proximity sensor, an illuminance sensor, an acceleration sensor, a magnetic sensor, a gyro sensor, an inertial sensor, an RGB sensor, an IR sensor, a fingerprint recognition sensor, an ultrasonic sensor, an optical sensor, a microphone, a lidar, and a radar. - The
output unit 150 may generate an output related to a visual sense, an auditory sense, or a haptic sense. - At this time, the
output unit 150 may include a display unit for outputting time information, a speaker for outputting auditory information, and a haptic module for outputting haptic information. - The
memory 170 may store data that supports various functions of theAI apparatus 100. For example, thememory 170 may store input data acquired by theinput unit 120, training data, a learning model, a learning history, and the like. - The
processor 180 may determine at least one executable operation of theAI apparatus 100 based on information determined or generated by using a data analysis algorithm or a machine learning algorithm. Theprocessor 180 may control the components of theAI apparatus 100 to execute the determined operation. - To this end, the
processor 180 may request, search, receive, or utilize data of the learningprocessor 130 or thememory 170. Theprocessor 180 may control the components of theAI apparatus 100 to execute the predicted operation or the operation determined to be desirable among the at least one executable operation. - When the connection of an external device is required to perform the determined operation, the
processor 180 may generate a control signal for controlling the external device and may transmit the generated control signal to the external device. - The
processor 180 may acquire intention information for the user input and may determine the user's requirements based on the acquired intention information. - The
processor 180 may acquire the intention information corresponding to the user input by using at least one of a speech to text (STT) engine for converting speech input into a text string or a natural language processing (NLP) engine for acquiring intention information of a natural language. - At least one of the STT engine or the NLP engine may be configured as an artificial neural network, at least part of which is learned according to the machine learning algorithm. At least one of the STT engine or the NLP engine may be learned by the learning
processor 130, may be learned by the learningprocessor 240 of theAI server 200, or may be learned by their distributed processing. - The
processor 180 may collect history information including the operation contents of theAI apparatus 100 or the user's feedback on the operation and may store the collected history information in thememory 170 or thelearning processor 130 or transmit the collected history information to the external device such as theAI server 200. The collected history information may be used to update the learning model. - The
processor 180 may control at least part of the components ofAI apparatus 100 so as to drive an application program stored inmemory 170. Furthermore, theprocessor 180 may operate two or more of the components included in theAI apparatus 100 in combination so as to drive the application program. -
FIG. 2 is a block diagram illustrating anAI server 200 according to an embodiment of the present invention. - Referring to
FIG. 2 , theAI server 200 may refer to a device that learns an artificial neural network by using a machine learning algorithm or uses a learned artificial neural network. TheAI server 200 may include a plurality of servers to perform distributed processing, or may be defined as a 5G network. At this time, theAI server 200 may be included as a partial configuration of theAI apparatus 100, and may perform at least part of the AI processing together. - The
AI server 200 may include acommunication unit 210, amemory 230, a learningprocessor 240, aprocessor 260, and the like. - The
communication unit 210 can transmit and receive data to and from an external device such as theAI apparatus 100. - The
memory 230 may include amodel storage unit 231. Themodel storage unit 231 may store a learning or learned model (or an artificialneural network 231 a) through the learningprocessor 240. - The learning
processor 240 may learn the artificialneural network 231 a by using the training data. The learning model may be used in a state of being mounted on theAI server 200 of the artificial neural network, or may be used in a state of being mounted on an external device such as theAI apparatus 100. - The learning model may be implemented in hardware, software, or a combination of hardware and software. If all or part of the learning models are implemented in software, one or more instructions that constitute the learning model may be stored in
memory 230. - The
processor 260 may infer the result value for new input data by using the learning model and may generate a response or a control command based on the inferred result value. -
FIG. 3 is a diagram illustrating anAI system 1 according to an embodiment of the present invention. - Referring to
FIG. 3 , in theAI system 1, at least one of anAI server 200, arobot 100 a, a self-drivingvehicle 100 b, anXR device 100 c, asmartphone 100 d, or ahome appliance 100 e is connected to acloud network 10. Therobot 100 a, the self-drivingvehicle 100 b, theXR device 100 c, thesmartphone 100 d, or thehome appliance 100 e, to which the AI technology is applied, may be referred to asAI apparatuses 100 a to 100 e. - The
cloud network 10 may refer to a network that forms part of a cloud computing infrastructure or exists in a cloud computing infrastructure. Thecloud network 10 may be configured by using a 3G network, a 4G or LTE network, or a 5G network. - That is, the
devices 100 a to 100 e and 200 configuring theAI system 1 may be connected to each other through thecloud network 10. In particular, each of thedevices 100 a to 100 e and 200 may communicate with each other through a base station, but may directly communicate with each other without using a base station. - The
AI server 200 may include a server that performs AI processing and a server that performs operations on big data. - The
AI server 200 may be connected to at least one of the AI apparatuses constituting theAI system 1, that is, therobot 100 a, the self-drivingvehicle 100 b, theXR device 100 c, thesmartphone 100 d, or thehome appliance 100 e through thecloud network 10, and may assist at least part of AI processing of theconnected AI apparatuses 100 a to 100 e. - At this time, the
AI server 200 may learn the artificial neural network according to the machine learning algorithm instead of theAI apparatuses 100 a to 100 e, and may directly store the learning model or transmit the learning model to theAI apparatuses 100 a to 100 e. - At this time, the
AI server 200 may receive input data from theAI apparatuses 100 a to 100 e, may infer the result value for the received input data by using the learning model, may generate a response or a control command based on the inferred result value, and may transmit the response or the control command to theAI apparatuses 100 a to 100 e. - Alternatively, the
AI apparatuses 100 a to 100 e may infer the result value for the input data by directly using the learning model, and may generate the response or the control command based on the inference result. - Hereinafter, various embodiments of the
AI apparatuses 100 a to 100 e to which the above-described technology is applied will be described. The AI apparatuses 100 a to 100 e illustrated inFIG. 3 may be regarded as a specific embodiment of theAI apparatus 100 illustrated inFIG. 1 . - <AI+Robot>
- The
robot 100 a, to which the AI technology is applied, may be implemented as a guide robot, a carrying robot, a cleaning robot, a wearable robot, an entertainment robot, a pet robot, an unmanned flying robot, or the like. - The
robot 100 a may include a robot control module for controlling the operation, and the robot control module may refer to a software module or a chip implementing the software module by hardware. - The
robot 100 a may acquire state information about therobot 100 a by using sensor information acquired from various kinds of sensors, may detect (recognize) surrounding environment and objects, may generate map data, may determine the route and the travel plan, may determine the response to user interaction, or may determine the operation. - The
robot 100 a may use the sensor information acquired from at least one sensor among the lidar, the radar, and the camera so as to determine the travel route and the travel plan. - The
robot 100 a may perform the above-described operations by using the learning model composed of at least one artificial neural network. For example, therobot 100 a may recognize the surrounding environment and the objects by using the learning model, and may determine the operation by using the recognized surrounding information or object information. The learning model may be learned directly from therobot 100 a or may be learned from an external device such as theAI server 200. - At this time, the
robot 100 a may perform the operation by generating the result by directly using the learning model, but the sensor information may be transmitted to the external device such as theAI server 200 and the generated result may be received to perform the operation. - The
robot 100 a may use at least one of the map data, the object information detected from the sensor information, or the object information acquired from the external device to determine the travel route and the travel plan, and may control the driving unit such that therobot 100 a travels along the determined travel route and travel plan. - The map data may include object identification information about various objects arranged in the space in which the
robot 100 a moves. For example, the map data may include object identification information about fixed objects such as walls and doors and movable objects such as pollen and desks. The object identification information may include a name, a type, a distance, and a position. - In addition, the
robot 100 a may perform the operation or travel by controlling the driving unit based on the control/interaction of the user. At this time, therobot 100 a may acquire the intention information of the interaction due to the user's operation or speech utterance, and may determine the response based on the acquired intention information, and may perform the operation. - <AI+Self-Driving>
- The self-driving
vehicle 100 b, to which the AI technology is applied, may be implemented as a mobile robot, a vehicle, an unmanned flying vehicle, or the like. - The self-driving
vehicle 100 b may include a self-driving control module for controlling a self-driving function, and the self-driving control module may refer to a software module or a chip implementing the software module by hardware. The self-driving control module may be included in the self-drivingvehicle 100 b as a component thereof, but may be implemented with separate hardware and connected to the outside of the self-drivingvehicle 100 b. - The self-driving
vehicle 100 b may acquire state information about the self-drivingvehicle 100 b by using sensor information acquired from various kinds of sensors, may detect (recognize) surrounding environment and objects, may generate map data, may determine the route and the travel plan, or may determine the operation. - Like the
robot 100 a, the self-drivingvehicle 100 b may use the sensor information acquired from at least one sensor among the lidar, the radar, or the camera so as to determine the travel route and the travel plan. - In particular, the self-driving
vehicle 100 b may recognize the environment or objects for an area covered by a field of view or an area over a certain distance by receiving the sensor information from external devices, or may receive directly recognized information from the external devices. - The self-driving
vehicle 100 b may perform the above-described operations by using the learning model composed of at least one artificial neural network. For example, the self-drivingvehicle 100 b may recognize the surrounding environment and the objects by using the learning model, and may determine the traveling movement line by using the recognized surrounding information or object information. The learning model may be learned directly from the self-drivingvehicle 100 a or may be learned from an external device such as theAI server 200. - At this time, the self-driving
vehicle 100 b may perform the operation by generating the result by directly using the learning model, but the sensor information may be transmitted to the external device such as theAI server 200 and the generated result may be received to perform the operation. - The self-driving
vehicle 100 b may use at least one of the map data, the object information detected from the sensor information, or the object information acquired from the external device to determine the travel route and the travel plan, and may control the driving unit such that the self-drivingvehicle 100 b travels along the determined travel route and travel plan. - The map data may include object identification information about various objects arranged in the space (for example, road) in which the self-driving
vehicle 100 b travels. For example, the map data may include object identification information about fixed objects such as street lamps, rocks, and buildings and movable objects such as vehicles and pedestrians. The object identification information may include a name, a type, a distance, and a position. - In addition, the self-driving
vehicle 100 b may perform the operation or travel by controlling the driving unit based on the control/interaction of the user. At this time, the self-drivingvehicle 100 b may acquire the intention information of the interaction due to the user's operation or speech utterance, and may determine the response based on the acquired intention information, and may perform the operation. - <AI+XR>
- The
XR device 100 c, to which the AI technology is applied, may be implemented by a head-mount display (HMD), a head-up display (HUD) provided in the vehicle, a television, a mobile phone, a smartphone, a computer, a wearable device, a home appliance, a digital signage, a vehicle, a fixed robot, a mobile robot, or the like. - The
XR device 100 c may analyzes three-dimensional point cloud data or image data acquired from various sensors or the external devices, generate position data and attribute data for the three-dimensional points, acquire information about the surrounding space or the real object, and render to output the XR object to be output. For example, theXR device 100 c may output an XR object including the additional information about the recognized object in correspondence to the recognized object. - The
XR device 100 c may perform the above-described operations by using the learning model composed of at least one artificial neural network. For example, theXR device 100 c may recognize the real object from the three-dimensional point cloud data or the image data by using the learning model, and may provide information corresponding to the recognized real object. The learning model may be directly learned from theXR device 100 c, or may be learned from the external device such as theAI server 200. - At this time, the
XR device 100 c may perform the operation by generating the result by directly using the learning model, but the sensor information may be transmitted to the external device such as theAI server 200 and the generated result may be received to perform the operation. - <AI+Robot+Self-Driving>
- The
robot 100 a, to which the AI technology and the self-driving technology are applied, may be implemented as a guide robot, a carrying robot, a cleaning robot, a wearable robot, an entertainment robot, a pet robot, an unmanned flying robot, or the like. - The
robot 100 a, to which the AI technology and the self-driving technology are applied, may refer to the robot itself having the self-driving function or therobot 100 a interacting with the self-drivingvehicle 100 b. - The
robot 100 a having the self-driving function may collectively refer to a device that moves for itself along the given movement line without the user's control or moves for itself by determining the movement line by itself. - The
robot 100 a and the self-drivingvehicle 100 b having the self-driving function may use a common sensing method so as to determine at least one of the travel route or the travel plan. For example, therobot 100 a and the self-drivingvehicle 100 b having the self-driving function may determine at least one of the travel route or the travel plan by using the information sensed through the lidar, the radar, and the camera. - The
robot 100 a that interacts with the self-drivingvehicle 100 b exists separately from the self-drivingvehicle 100 b and may perform operations interworking with the self-driving function of the self-drivingvehicle 100 b or interworking with the user who rides on the self-drivingvehicle 100 b. - At this time, the
robot 100 a interacting with the self-drivingvehicle 100 b may control or assist the self-driving function of the self-drivingvehicle 100 b by acquiring sensor information on behalf of the self-drivingvehicle 100 b and providing the sensor information to the self-drivingvehicle 100 b, or by acquiring sensor information, generating environment information or object information, and providing the information to the self-drivingvehicle 100 b. - Alternatively, the
robot 100 a interacting with the self-drivingvehicle 100 b may monitor the user boarding the self-drivingvehicle 100 b, or may control the function of the self-drivingvehicle 100 b through the interaction with the user. For example, when it is determined that the driver is in a drowsy state, therobot 100 a may activate the self-driving function of the self-drivingvehicle 100 b or assist the control of the driving unit of the self-drivingvehicle 100 b. The function of the self-drivingvehicle 100 b controlled by therobot 100 a may include not only the self-driving function but also the function provided by the navigation system or the audio system provided in the self-drivingvehicle 100 b. - Alternatively, the
robot 100 a that interacts with the self-drivingvehicle 100 b may provide information or assist the function to the self-drivingvehicle 100 b outside the self-drivingvehicle 100 b. For example, therobot 100 a may provide traffic information including signal information and the like, such as a smart signal, to the self-drivingvehicle 100 b, and automatically connect an electric charger to a charging port by interacting with the self-drivingvehicle 100 b like an automatic electric charger of an electric vehicle. - <AI+Robot+XR>
- The
robot 100 a, to which the AI technology and the XR technology are applied, may be implemented as a guide robot, a carrying robot, a cleaning robot, a wearable robot, an entertainment robot, a pet robot, an unmanned flying robot, a drone, or the like. - The
robot 100 a, to which the XR technology is applied, may refer to a robot that is subjected to control/interaction in an XR image. In this case, therobot 100 a may be separated from theXR device 100 c and interwork with each other. - When the
robot 100 a, which is subjected to control/interaction in the XR image, may acquire the sensor information from the sensors including the camera, therobot 100 a or theXR device 100 c may generate the XR image based on the sensor information, and theXR device 100 c may output the generated XR image. Therobot 100 a may operate based on the control signal input through theXR device 100 c or the user's interaction. - For example, the user can confirm the XR image corresponding to the time point of the
robot 100 a interworking remotely through the external device such as theXR device 100 c, adjust the self-driving travel path of therobot 100 a through interaction, control the operation or driving, or confirm the information about the surrounding object. - <AI+Self-Driving+XR>
- The self-driving
vehicle 100 b, to which the AI technology and the XR technology are applied, may be implemented as a mobile robot, a vehicle, an unmanned flying vehicle, or the like. - The self-driving
driving vehicle 100 b, to which the XR technology is applied, may refer to a self-driving vehicle having a means for providing an XR image or a self-driving vehicle that is subjected to control/interaction in an XR image. Particularly, the self-drivingvehicle 100 b that is subjected to control/interaction in the XR image may be distinguished from theXR device 100 c and interwork with each other. - The self-driving
vehicle 100 b having the means for providing the XR image may acquire the sensor information from the sensors including the camera and output the generated XR image based on the acquired sensor information. For example, the self-drivingvehicle 100 b may include an HUD to output an XR image, thereby providing a passenger with a real object or an XR object corresponding to an object in the screen. - At this time, when the XR object is output to the HUD, at least part of the XR object may be outputted so as to overlap the actual object to which the passenger's gaze is directed. Meanwhile, when the XR object is output to the display provided in the self-driving
vehicle 100 b, at least part of the XR object may be output so as to overlap the object in the screen. For example, the self-drivingvehicle 100 b may output XR objects corresponding to objects such as a lane, another vehicle, a traffic light, a traffic sign, a two-wheeled vehicle, a pedestrian, a building, and the like. - When the self-driving
vehicle 100 b, which is subjected to control/interaction in the XR image, may acquire the sensor information from the sensors including the camera, the self-drivingvehicle 100 b or theXR device 100 c may generate the XR image based on the sensor information, and theXR device 100 c may output the generated XR image. The self-drivingvehicle 100 b may operate based on the control signal input through the external device such as theXR device 100 c or the user's interaction. -
FIG. 4 is a block diagram illustrating an AI apparatus according to an embodiment of the present invention. - The redundant repeat of
FIG. 1 will be omitted below. - Referring to
FIG. 4 , theinput unit 120 may include acamera 121 for image signal input, amicrophone 122 for receiving audio signal input, and auser input unit 123 for receiving information from a user. - Voice data or image data collected by the
input unit 120 are analyzed and processed as a user's control command. - Then, the
input unit 120 is used for inputting image information (or signal), audio information (or signal), data, or information inputted from a user and theAI apparatus 100 may include at least onecamera 121 in order for inputting image information. - The
camera 121 processes image frames such as a still image or a video obtained by an image sensor in a video call mode or a capturing mode. The processed image frame may be displayed on thedisplay unit 151 or stored in thememory 170. - The
microphone 122 processes external sound signals as electrical voice data. The processed voice data may be utilized variously according to a function (or an application program being executed) being performed in theAI apparatus 100. Moreover, various noise canceling algorithms for removing noise occurring during the reception of external sound signals may be implemented in themicrophone 122. - The
user input unit 123 is to receive information from a user and when information is inputted through theuser input unit 123, theprocessor 180 may control an operation of theAI apparatus 100 to correspond to the inputted information. - The
user input unit 123 may include a mechanical input means (or a mechanical key, for example, a button, a dome switch, a jog wheel, and a jog switch at the front, back or side of the AI apparatus 100) and a touch type input means. As one example, a touch type input means may include a virtual key, a soft key, or a visual key, which is displayed on a touch screen through software processing or may include a touch key disposed at a portion other than the touch screen. - A
sensing unit 140 may be called a sensor unit. - The
output unit 150 may include at least one of adisplay unit 151, asound output module 152, ahaptic module 153, or anoptical output module 154. - The
display unit 151 may display (output) information processed in theAI apparatus 100. For example, thedisplay unit 151 may display execution screen information of an application program running on theAI apparatus 100 or user interface (UI) and graphic user interface (GUI) information according to such execution screen information. - The
display unit 151 may be formed with a mutual layer structure with a touch sensor or formed integrally, so that a touch screen may be implemented. Such a touch screen may serve as theuser input unit 123 providing an input interface between theAI apparatus 100 and a user, and an output interface between theAI apparatus 100 and a user at the same time. - The
sound output module 152 may output audio data received from thewireless communication unit 110 or stored in thememory 170 in a call signal reception or call mode, a recording mode, a voice recognition mode, or a broadcast reception mode. - The
sound output module 152 may include a receiver, a speaker, and a buzzer. - The
haptic module 153 generates various haptic effects that a user can feel. A representative example of a haptic effect that thehaptic module 153 generates is vibration. - The
optical output module 154 outputs a signal for notifying event occurrence by using light of a light source of theAI apparatus 100. An example of an event occurring in theAI apparatus 100 includes message reception, call signal reception, missed calls, alarm, schedule notification, e-mail reception, and information reception through an application. -
FIG. 5 is a flowchart illustrating a method for recognizing speech of a user in consideration of user's application usage according to an embodiment of the present invention. - Referring to
FIG. 5 , theprocessor 180 of theAI apparatus 100 receives a sound signal for the speech of the user (S501). - The sound signal of the user's speech may be received via the
microphone 121 of theAI apparatus 100, but may also be received from an external device (not shown) via thecommunication unit 110. - The sound signal may be an audio file in a pulse code modulation (PCM) format.
- Then, the
processor 180 of theAI apparatus 100 recognizes the speech using a language model (LM) (S503). - Here, the recognizing of the speech may mean generating a text string corresponding to the sound signal.
- The
processor 180 may recognize the input sound signal on a phoneme basis using an acoustic model (AM) and may determine a word corresponding to the input sound signal based on the recognition result on a phoneme basis. Theprocessor 180 may generate a word vector on a phoneme basis using the recognition result on a phoneme basis and determine at least one word corresponding to the input sound signal using the generated word vector on a phoneme basis and the language model. - The at least one word determined by the
processor 180 may include a single word or a word sequence composed of a plurality of words. - The language model may refer to a model that represents a probability distribution for the word sequence. That is, when a specific word sequence is given, the language model may output a probability that the word sequence appears. Thus, the
processor 180 may use the language model to determine candidate words suitable for a next word when a specific word sequence is given or to calculate a probability that each candidate word appears. - For example, it is assumed that the
processor 180 recognizes phonemes of a specific recognition target word as [‘n’, ‘i’, ‘g’, ‘h’, ‘t’] using the acoustic model. Theprocessor 180 may generate a word vector on a phoneme basis corresponding to the [‘n’, ‘i’, ‘g’, ‘h’, ‘t’] and determine the recognition target word using the generated word vector and the language model. When the language model is used, and when it is determined that the recognition target word is more likely to be a “knight” than a “night” in view of the word sequence recognized so far, theprocessor 180 may determine the target word to be the “knight” rather than the “night”, unlike the result of the acoustic model. - Although the above example describes a situation where only a single word is determined, the present invention is not limited thereto. That is, the
processor 180 may determine a single word using word vectors on a phoneme basis estimated to be a plurality of words, or conversely, determine a plurality of words using a word vector on a phoneme basis estimated to be a single word. - For example, even when the recognition result on a phoneme basis from the acoustic model indicates a single word “foreign”, the recognition target word may be determined as a plurality of words “for him”. On the contrary, even when the recognition result on a phoneme basis from the acoustic model indicates the plurality of words “for him”, the recognition target word may be determined as the single word “foreign” based on the determination result of the language model.
- In addition, the
processor 180 may recognize a word based on the recognition result on a phoneme basis of the acoustic model and calculate an LM score of the language model for the recognized word. Further, when the LM score of the recognized word is below a certain level, theprocessor 180 may find a word similar to the recognized word but has a higher LM score from a lexicon and determine the found word to be a recognition result. - That is, the
processor 180 may recognize the word more accurately using the word vector on a phoneme basis and the language model for the word, which is the current recognition target, together. Thus, theprocessor 180 may more accurately recognize the speech of the user in general. - Here, the language model may be learned as a personalized model classified for each user, and the
processor 180 may recognize the user when the sound signal is received and recognize the speech using the personalized language model corresponding to the recognized user. - Each of the acoustic model or the language model described above may be a model learned using a machine learning algorithm or a deep learning algorithm, and may be configured as an artificial neural network. Learning of the acoustic model or the language model may be performed in the
learning processor 130 of theAI apparatus 100 or in thelearning processor 240 of theAI server 200. - The acoustic model or the language model may be stored in the
memory 170 of theAI apparatus 100 or may be stored in thememory 230 of theAI server 200. - Then, the
processor 180 of theAI apparatus 100 determines the intention of the user based on the speech recognition result (S505). - The determining of the intention of the user may mean recognizing the intention of the user. Further, the determining of the intention corresponding to the user's speech may mean determining the intention of the natural language corresponding to the speech. Further, the intention of the user may mean an intention corresponding to the user's speech.
- The
processor 180 may determine the intention of the user by generating the intention information based on the speech recognition result. - Here, the
processor 180 may determine the intention of the user using an intention classifier of a natural language understanding (NLU) technique. - The intention classifier may refer to a model which projects an intention of the input speech recognition result onto a vector space, when the speech recognition result is inputted.
- Here, a cluster may be formed for each intention in the vector space resulted from an output of the intention classifier. That is, there may be a plurality of intention clusters that may be classified by the intention classifier in the vector space resulted from the output of the intention classifier.
- Here, when the speech recognition result is input to the intention classifier, the more the intention of the input speech recognition result and a specific intention cluster are similar, the intention classifier may project the input speech recognition result to a location adjacent to the corresponding intention cluster. Therefore, a distance between the projected location and each intention cluster means a similarity between the intention of the input speech recognition result and each intention cluster. Thus, a short distance may mean a high degree of the similarity.
- Here, the intention classifier may be an artificial neural network-based model learned using the machine learning algorithm or the deep learning algorithm.
- The
processor 180 of theAI apparatus 100 determines whether the intention recognition is successful (S507). - Whether the intention recognition is successful may mean whether accuracy or reliability of the intention recognition exceeds a preset first reference value.
- Here, the
processor 180 may calculate the reliability of the intention recognition using the distance from the location to which the speech recognition result input in the vector space is projected by the intention classifier to each of the intention clusters. - Here, the
processor 180 may calculate the reliability of the intention recognition using a distance from the location to which the speech recognition result input in the vector space is projected to the nearest intention cluster. This is because an intention corresponding to the intention cluster nearest to the location to which the input speech recognition result is projected may be determined to be the intention of the input speech recognition result. - Hereinafter, the distance from the location to which the input speech recognition result is projected to the nearest intention cluster may be referred to as a first cluster distance. Further, a distance from the location to which the input speech recognition result is projected to an intention cluster distant at an n-th spacing may be referred to as an n-th cluster distance. Similarly, the intention cluster distant at an n-th spacing from the location to which the input speech recognition result is projected may be referred to as an n-th cluster.
- For example, as the first cluster distance is smaller, the
processor 180 may calculate higher reliability of the intention recognition. Further, theprocessor 180 may determine whether the intention recognition is successful based on whether the calculated reliability exceeds the first reference value. - Here, the
processor 180 may calculate the reliability of the intention recognition using both the first cluster distance and the second cluster distance. - For example, as the first cluster distance is smaller and the difference between the first cluster distance and the second cluster distance is larger, the
processor 180 may calculate higher reliability of the intention recognition. Further, theprocessor 180 may determine whether the intention recognition is successful based on whether the calculated reliability exceeds the first reference value. This is because it may be seen that the difference between the first cluster distance and the second cluster distance indicates how assuredly the intention analyzer determined the intention of the input speech recognition result to be an intention that corresponds to the first cluster. - Alternatively, the
processor 180 may determine whether the intention recognition is successful by determining whether the first cluster distance is smaller than a preset second reference value without explicitly calculating the reliability of the intention recognition. - Alternatively, the
processor 180 may determine whether the intention recognition is successful by determining whether the first cluster distance is smaller than a preset third reference value and whether a difference between the first cluster distance and the second distance is larger than a preset fourth reference value. - If it is determined in S507 that the intention of the user is successfully recognized, the
processor 180 of theAI apparatus 100 performs an operation corresponding to the determined intention of the user (S509). - For example, when the user requests specific information, the
processor 180 may output corresponding information via the speaker or the display unit. - If it is determined in S507 that the user's intention recognition is failed, the
processor 180 of theAI apparatus 100 obtains a user's application usage log (S511). - The application usage log may include not only an application usage log in the
AI apparatus 100 but also an application usage log in an external device capable of communicating with theAI apparatus 100. - Further, the application usage log may include a type of an executed application, a content of user's manipulation for the executed application, and the like. Further, the content of the user's manipulation for the executed application may include a search history of the user, a browse history of the search result of the user, and the like.
- Here, the
processor 180 may obtain the application usage log before and after the user's utterance. This is because a pattern of the usage of the application before and after the user's utterance may be expected to be highly related to utterance content. - For example, when the user is running a weather application or searching weather on the Internet in a situation where the user has asked a specific question but the
AI apparatus 100 did not grasp the intention of the user, the user's question is likely to be a question about the weather. - Similarly, when the user has asked a specific question while using a movie ticket booking application but the
AI apparatus 100 did not grasp the intention of the user, and when the recognized word is similar to a movie title displayed in the movie ticket booking application, the user's question is likely to be a question asking for information about the corresponding movie. - Therefore, the application usage log before and after the user's utterance helps to grasp the content of the user's utterance and the intention of the utterance, and may be used as implicit feedback. Here, the
processor 180 may obtain application usage log for a predetermined period from an utterance time point. - For example, the predetermined period may be determined to be a period including the utterance time point, such as a time period from 30 seconds before the utterance time point to 1 minute after the utterance time or may be determined to be a period from the utterance time point.
- Here, the
processor 180 may determine whether an permission to obtain the application usage log exists. Further, when the obtaining permission does not exist, theprocessor 180 may output a notification requesting the obtaining permission. - Then, the
processor 180 of theAI apparatus 100 updates the language model using the obtained user's application usage log (S513). - The
processor 180 may set a word mapping relationship or adjust a weighted value, so that the language model corrects incorrectly recognized words based on the application usage log. - For example, it is assumed that the
processor 180 recognized the speech of the user using the language model and the recognition result is “two dies weather” and it is assumed that the user runs the weather application or searches “today's weather” on the Internet. In this case, theprocessor 180 may map the “two dies” to a “today's” for the language model or increase a weighted value of the “today's”, so that the language model may recognize the “today's” rather than the “two dies” for the same speech later based on the application usage log. - In particular, the
processor 180 may update the language model by mapping incorrectly recognized words to popular words in consideration of the user's dialect, intonation, and unusual word habits. - Here, the setting of the mapping relationship of the words and the adjusting of the weighted value in the language model may mean word embedding learning.
- The
processor 180 may compare the recognition result of the language model with the application usage log to determine whether the application usage log is related to the utterance of the user. Further, the language model may only be updated when it is determined that the application usage log is relevant to the utterance of the user. - Here, the
processor 180 may extract a keyword to be used for the language model update from the application usage log based on the recognition result of the language model and the application usage log. Further, theprocessor 180 may find a word (or keyword) corresponding to the extracted keyword among the recognition results from the language model, map the corresponding word found to the extracted keyword for the language model, or increase a weighted value of the extracted keyword. In addition, theprocessor 180 may increase a weighted value of each of words having pronunciation similar to that of the extracted keyword. - The
processor 180 may update the language model for each user. - The updated language model may be used to recognize future speech of the user.
- As described above, the language model may be used to recognize the user's speech, and may determine a word corresponding to the speech from the recognition result on a phoneme basis. However, the speech recognition performance of the general language model may deteriorate due to poor pronunciation, habit, or dialect of the user. However, a language model specific to an utterance feature may be constructed for each user by obtaining the implicit feedback for a situation where the speech recognition fails based on the user's application usage log and updating the language model for each user based on the implicit feedback.
-
FIG. 6 is a diagram illustrating a process of determining whether user's intention recognition is successful, according to an embodiment of the present invention. - Referring to
FIG. 6 , theprocessor 180 of theAI apparatus 100 may determine whether the user's intention recognition is successful from aspeech recognition result 610 using anintention classifier 620 of the natural language understanding technique. - As described above, when the speech recognition result is input, the
intention classifier 620 may project the intention from thespeech recognition result 610 input onto avector space 630. Further, a plurality of intention clusters that may be classified by theintention classifier 620 may exist in thevector space 630. - As shown in
FIG. 6 , when theintention classifier 620 projects the speech recognition result 610 alocation 631 away from all the intention clusters in thevector space 630, theprocessor 180 may calculate reliability of the intention recognition at a low level and thus determine that the recognition of the intention has failed. -
FIG. 7 is a flowchart illustrating an example of updating a language model (S513) shown inFIG. 5 . - Referring to
FIG. 7 , theprocessor 180 of theAI apparatus 100 calculates correlation between the speech recognition result and the application usage log (S701). - Even though the
AI apparatus 100 does not correctly determine the intention corresponding to the user's speech, the user may show a pattern of the usage of the application, which is not related to the speech. - Accordingly, the
processor 180 may calculate the correlation between the user's application usage log and the speech recognition result to determine whether the application usage log may be used to update the language model. - Here, the correlation between the speech recognition result and the application usage log may mean a similarity between the speech recognition result and the application usage log.
- Here, the
processor 180 may determine a similarity of pronunciation and meaning between words or keywords included in the speech recognition result and words or keywords included in the application usage log, and may calculate the correlation between the speech recognition result and the application usage log based on the similarity of the pronunciation and meaning. - The words or keywords included in the application usage log may include a keyword indicating a type and a name of the application included in the usage log, a keyword indicating information contained in the application, and the like.
- Further, the
processor 180 of theAI apparatus 100 determines whether the calculated correlation is equal to or above a preset reference value (S703). - When it is determined in S705 that the calculated association is equal to or above the preset reference value, the
processor 180 of theAI apparatus 100 updates the language model (S705). - In calculating the correlation between the speech recognition result and the application usage log in S701, the
processor 180 may determine a mapping relationship between the keyword of the speech recognition result and the keyword of the application usage log, and then theprocessor 180 may update the language model based on the determined mapping relationship. - As a result of the determination in S705, a procedure is terminated without updating the language model.
-
FIGS. 8 and 9 are diagrams illustrating a method for recognizing speech of a user according to an embodiment of the present invention. - Referring to
FIG. 8 , theuser 811 may givespeech utterance 812 “Today's weather” to theAI apparatus 821 to ask about today's weather. However, the processor of theAI apparatus 821 may incorrectly recognize the user'sspeech utterance 812 as “Two dies weather” (822). Thus, the processor may fail to grasp the intention of the user and suggest the user to search theweb 823. - For example, when the
user 811 is an English speaker or an Australian speaker, a pronunciation of the “Today” becomes similar to that of “Todai”. Therefore, the “Today's” may be recognized as “Two dies” when an acoustic model and a language model based on English pronunciation of American speaker are used. - The processor of the
AI apparatus 821 may use the intention classifier to determine that thespeech recognition result 822 in thevector space 831 is spaced apart from all the intention clusters. Thus, the processor may determine that the intention recognition has failed. - When the
user 811searches 842 for “Today's weather” on a search engine using the terminal 841, the processor of theAI apparatus 821 may map 851 the “Two dies” to “Today's” using this application usage log, and update the language model based on this mapping relationship. Here, the terminal 841 may be the same device as theAI apparatus 821, but may be a separate device that is different from theAI apparatus 821. -
FIG. 9 illustrates a process of recognizing speech of a user after a language model is updated according toFIG. 8 . - Referring to
FIG. 9 , when auser 811 gives thespeech utterance 812 of the “Today's weather”, which is the same as above, even when the acoustic model of theAI apparatus 821 recognizes pronunciation on a phoneme basis as before, the updated language model may recognize 922 the user'sspeech utterance 812 as “Today's weather” instead of “Two dies weather”. Accordingly, theAI apparatus 821 may provide the user withcurrent weather information 923. - According to an embodiment of the present invention, the above-described method may be implemented as a processor-readable code in a medium where a program is recorded. Examples of a processor-readable medium may include read-only memory (ROM), random access memory (RAM), CD-ROM, a magnetic tape, a floppy disk, and an optical data storage device.
Claims (14)
Applications Claiming Priority (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
KR10-2019-0088516 | 2019-07-22 | ||
KR1020190088516A KR102339085B1 (en) | 2019-07-22 | 2019-07-22 | Artificial intelligence apparatus for recognizing speech of user in consideration of user's application usage log and method for the same |
Publications (2)
Publication Number | Publication Date |
---|---|
US20190371297A1 true US20190371297A1 (en) | 2019-12-05 |
US11398222B2 US11398222B2 (en) | 2022-07-26 |
Family
ID=67808083
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
US16/539,773 Active 2040-01-20 US11398222B2 (en) | 2019-07-22 | 2019-08-13 | Artificial intelligence apparatus and method for recognizing speech of user in consideration of user's application usage log |
Country Status (2)
Country | Link |
---|---|
US (1) | US11398222B2 (en) |
KR (1) | KR102339085B1 (en) |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111369981A (en) * | 2020-03-02 | 2020-07-03 | 北京远鉴信息技术有限公司 | Dialect region identification method and device, electronic equipment and storage medium |
WO2021142442A1 (en) * | 2020-01-10 | 2021-07-15 | Optimus Ride, Inc. | Communication system and method |
Families Citing this family (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110544470B (en) * | 2019-09-11 | 2022-03-29 | 拉扎斯网络科技(上海)有限公司 | Voice recognition method and device, readable storage medium and electronic equipment |
CN110784384B (en) * | 2019-10-16 | 2021-11-02 | 杭州九阳小家电有限公司 | Household appliance voice technology generation method and intelligent household appliance |
KR102518448B1 (en) * | 2021-05-14 | 2023-04-06 | (주)와이즈에이아이 | System and method for providing automatic response call service based on ai chatbot |
Family Cites Families (15)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US6125347A (en) * | 1993-09-29 | 2000-09-26 | L&H Applications Usa, Inc. | System for controlling multiple user application programs by spoken input |
US6952800B1 (en) * | 1999-09-03 | 2005-10-04 | Cisco Technology, Inc. | Arrangement for controlling and logging voice enabled web applications using extensible markup language documents |
JP4281543B2 (en) * | 2003-12-16 | 2009-06-17 | 日産自動車株式会社 | VEHICLE DRIVE OPERATION ASSISTANCE DEVICE AND VEHICLE HAVING VEHICLE DRIVE OPERATION ASSISTANCE DEVICE |
KR20140088836A (en) * | 2011-03-28 | 2014-07-11 | 암비엔즈 | Methods and systems for searching utilizing acoustical context |
US20130227352A1 (en) * | 2012-02-24 | 2013-08-29 | Commvault Systems, Inc. | Log monitoring |
US20140250147A1 (en) * | 2013-03-01 | 2014-09-04 | Quixey, Inc. | Generating Search Results Containing State Links to Applications |
US20150370787A1 (en) * | 2014-06-18 | 2015-12-24 | Microsoft Corporation | Session Context Modeling For Conversational Understanding Systems |
US20160071517A1 (en) * | 2014-09-09 | 2016-03-10 | Next It Corporation | Evaluating Conversation Data based on Risk Factors |
KR102386863B1 (en) | 2015-09-09 | 2022-04-13 | 삼성전자주식회사 | User-based language model generating apparatus, method and voice recognition apparatus |
KR102429583B1 (en) * | 2016-12-14 | 2022-08-05 | 삼성전자주식회사 | Electronic apparatus, method for providing guide ui of thereof, and non-transitory computer readable recording medium |
KR102502220B1 (en) * | 2016-12-20 | 2023-02-22 | 삼성전자주식회사 | Electronic apparatus, method for determining user utterance intention of thereof, and non-transitory computer readable recording medium |
KR20180074210A (en) * | 2016-12-23 | 2018-07-03 | 삼성전자주식회사 | Electronic device and voice recognition method of the electronic device |
WO2018182311A1 (en) * | 2017-03-28 | 2018-10-04 | Samsung Electronics Co., Ltd. | Method for operating speech recognition service, electronic device and system supporting the same |
KR102421745B1 (en) * | 2017-08-22 | 2022-07-19 | 삼성전자주식회사 | System and device for generating TTS model |
US11069346B2 (en) * | 2019-04-22 | 2021-07-20 | International Business Machines Corporation | Intent recognition model creation from randomized intent vector proximities |
-
2019
- 2019-07-22 KR KR1020190088516A patent/KR102339085B1/en active IP Right Grant
- 2019-08-13 US US16/539,773 patent/US11398222B2/en active Active
Cited By (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2021142442A1 (en) * | 2020-01-10 | 2021-07-15 | Optimus Ride, Inc. | Communication system and method |
US11866063B2 (en) | 2020-01-10 | 2024-01-09 | Magna Electronics Inc. | Communication system and method |
CN111369981A (en) * | 2020-03-02 | 2020-07-03 | 北京远鉴信息技术有限公司 | Dialect region identification method and device, electronic equipment and storage medium |
Also Published As
Publication number | Publication date |
---|---|
KR20190098089A (en) | 2019-08-21 |
KR102339085B1 (en) | 2021-12-14 |
US11398222B2 (en) | 2022-07-26 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
US11663516B2 (en) | Artificial intelligence apparatus and method for updating artificial intelligence model | |
US11302311B2 (en) | Artificial intelligence apparatus for recognizing speech of user using personalized language model and method for the same | |
US11200467B2 (en) | Artificial intelligence apparatus and method for recognizing object included in image data | |
US11495214B2 (en) | Artificial intelligence device for providing voice recognition service and method of operating the same | |
US11282522B2 (en) | Artificial intelligence apparatus and method for recognizing speech of user | |
US11164586B2 (en) | Artificial intelligence apparatus and method for recognizing utterance voice of user | |
US11295166B2 (en) | Artificial intelligence apparatus for generating training data for artificial intelligence model and method thereof | |
US11398222B2 (en) | Artificial intelligence apparatus and method for recognizing speech of user in consideration of user's application usage log | |
US11669781B2 (en) | Artificial intelligence server and method for updating artificial intelligence model by merging plurality of pieces of update information | |
US11568239B2 (en) | Artificial intelligence server and method for providing information to user | |
US10872438B2 (en) | Artificial intelligence device capable of being controlled according to user's gaze and method of operating the same | |
US11810575B2 (en) | Artificial intelligence robot for providing voice recognition function and method of operating the same | |
US11769508B2 (en) | Artificial intelligence apparatus | |
US20190392816A1 (en) | Speech processing device and speech processing method | |
US20220338281A1 (en) | Artificial intelligence device for providing device control function based on interworking between devices and method therefor | |
US11270700B2 (en) | Artificial intelligence device and method for recognizing speech with multiple languages | |
US11455529B2 (en) | Artificial intelligence server for controlling a plurality of robots based on guidance urgency | |
US20210334461A1 (en) | Artificial intelligence apparatus and method for generating named entity table | |
US11322134B2 (en) | Artificial intelligence device and operating method thereof | |
US11676012B2 (en) | Artificial intelligence server | |
US11423877B2 (en) | Robot for providing guidance service using artificial intelligence and method of operating the same | |
US20190377489A1 (en) | Artificial intelligence device for providing voice recognition service and method of operating the same | |
US11348585B2 (en) | Artificial intelligence apparatus | |
US20210118437A1 (en) | Artificial intelligence server | |
US11595227B2 (en) | Artificial intelligence device and operating method thereof |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
AS | Assignment |
Owner name: LG ELECTRONICS INC., KOREA, REPUBLIC OF Free format text: ASSIGNMENT OF ASSIGNORS INTEREST;ASSIGNORS:KIM, JAEHONG;KIM, BOSEOP;REEL/FRAME:050042/0766 Effective date: 20190731 |
|
FEPP | Fee payment procedure |
Free format text: ENTITY STATUS SET TO UNDISCOUNTED (ORIGINAL EVENT CODE: BIG.); ENTITY STATUS OF PATENT OWNER: LARGE ENTITY |
|
STPP | Information on status: patent application and granting procedure in general |
Free format text: DOCKETED NEW CASE - READY FOR EXAMINATION |
|
STPP | Information on status: patent application and granting procedure in general |
Free format text: NON FINAL ACTION MAILED |
|
STPP | Information on status: patent application and granting procedure in general |
Free format text: RESPONSE TO NON-FINAL OFFICE ACTION ENTERED AND FORWARDED TO EXAMINER |
|
STPP | Information on status: patent application and granting procedure in general |
Free format text: FINAL REJECTION MAILED |
|
STPP | Information on status: patent application and granting procedure in general |
Free format text: DOCKETED NEW CASE - READY FOR EXAMINATION |
|
STPP | Information on status: patent application and granting procedure in general |
Free format text: NON FINAL ACTION MAILED |
|
STPP | Information on status: patent application and granting procedure in general |
Free format text: RESPONSE TO NON-FINAL OFFICE ACTION ENTERED AND FORWARDED TO EXAMINER |
|
STPP | Information on status: patent application and granting procedure in general |
Free format text: NOTICE OF ALLOWANCE MAILED -- APPLICATION RECEIVED IN OFFICE OF PUBLICATIONS |
|
STCF | Information on status: patent grant |
Free format text: PATENTED CASE |